bioelectrochemical systems for energy ...―wastewater treatment process‖. australian provisional...
TRANSCRIPT
BIOELECTROCHEMICAL SYSTEMS FOR
ENERGY RECOVERY FROM WASTEWATER
KA YU CHENG
Bioelectrochemical Systems
for Energy Recovery from Wastewater
Ka Yu Cheng
BSc (Hons); M.Phil.
A thesis submitted in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
in
Environmental Engineering
Faculty of Sustainability, Environmental and Life Sciences
Murdoch University
WA, Australia
November 2009
i
Abstract
The global concerns of climate change and energy crisis have provoked research efforts
to develop energy-efficient alternatives to conventional activated sludge wastewater treatment
processes. Recently, bio-electrochemical systems (BES) such as microbial fuel cells (MFCs) and
microbial electrolysis cells (MECs) have emerged as a promising technology for simultaneous
energy recovery and wastewater treatment. These systems harness the capacity of
microorganisms for the catalysis of electrochemical reaction. In MFCs, chemical energy in the
form of organic compounds in wastewater is directly converted into electricity. While in MECs,
external electricity is provided to enable more valuable products (e.g. hydrogen) to form at the
cathode.
Over the past decade, our knowledge on BES is gaining momentum. However,
wastewater treatment using BES has not been successful on an industrial-scale. Many
technological bottlenecks still remain unresolved and our understanding of microbe-electrode
interactions in BES is incomplete. The overall aim of this thesis is to generate understanding that
will be helpful in the development of an energy recovering wastewater treatment process using
BESs. The scope of this thesis comprises two themes. The first theme is to study the
fundamental aspect of BESs with an emphasis on microbe-anode interactions, the reaction that
makes it possible to use organic waste substances as the electron donor for BES. The second
theme is to quantify rate limiting steps and to develop practical solutions to overcome the
established bottlenecks making room for the development of new technologies of BES for
wastewater treatment and related purposes.
Most of the experiments were conducted using a sub-liter scale two-chamber BES
equipped with a cation exchange membrane. An electrochemically active biofilm was
established at the anode (graphite granules) from an activated sludge inoculum. A synthetic
wastewater with acetate as the sole electron donor was used throughout the study. The cathodic
electron acceptor was either potassium ferricyanide (K3Fe(CN)6) or dissolved oxygen. No
chemical catalyst (e.g. platinum) was applied to the electrodes. In some experiments, the BES
was coupled with a potentiostat to precisely control the biofilm-electrode potential or to perform
voltammetric analysis.
The results showed that activated sludge bacteria could readily initiate a highly effective
anodophilic biofilm (415 W·m-3
after five days with a Fe(CN)63+
cathode), provided that factors
such as electrolyte pH, external resistance and cathodic oxidizing power were not limiting. From
the Coulombic efficiency of over 80% the microbial activity could be recorded by online
monitoring of the current. This allowed a detailed study of the affinity for the anode of biofilm.
In analogy to the well known Michaelis-Menten kinetics, a half-saturation anodic potential (kAP)
was established at which the microbial metabolic rate reached half its maximum rate. This kAP
value was about -455 mV (vs. Ag/AgCl) for our acetate-driven biofilm. A critical anodic
potential (APcrit.) of about -420 mV (vs. Ag/AgCl) was defined that characterizes both the
bacterial saturation by the electron-accepting system and the maximal MFC power output. This
information is useful for MFC modeling and optimization.
Although online process control is used for many bioprocesses it is not established for
MFC. A new approach was developed that enabled voltammetric studies of the biofilm-anode,
ii
without using potentiostats but by feedback controlling a variable external resistance of the
running MFC. This approach could perform the conventional cyclic voltammetry of a MFC
without interrupting its operation.
In MFC the anodic reaction is proton liberating while the cathodic reaction is proton
consuming. This leads to perhaps the major limitation in MFC operation known as a pH gradient.
This limitation was addressed by using a novel operational regime: the intermittent polarity
inversion. At electrode potential of -300 mV (vs. Ag/AgCl) the alternating supply of acetate and
oxygen to the biofilm resulted in the generation of an anodic and cathodic current of +240 and -
80 mA (+1500 and -500 A·m-3
), respectively. Since the anodic reaction is proton-liberating
while the cathodic reaction is proton-consuming, such operational regime prevented the
detrimental build-up of a pH gradient enabling a prolonged operation of the MFC without using
costly pH control methods (dosing of acid/base or chemical buffers).
The intermittent polarity inversion showed signs of the ―anodophilic bacteria‖ being able
to catalyze not only the anodic oxidation of acetate but also the cathodic oxygen reduction. The
presence of ―anodophilic bacteria‖ at the cathode could enable a 5-fold increase of power output
(from 5.6 to 27 W·m-3
). This is the first evidence that a BES biofilm can catalyze both the
forward and backward electron flow with a single electrode. Based on this finding, a novel
scalable, membrane-less BES configuration, termed rotatable bio-electrochemical contactor
(RBEC) has been developed.
Similar to rotating biological contactors (RBC), the RBEC consists of a cylindrical
water-holding vessel (ca. 3 L) which houses an array of carbon cloth coated discs (electrodes)
mounted onto a central horizontal rotatable shaft. Each disc consists of a water-immersed anodic
and an air-exposed cathodic half connected via a resistor. No ion-exchange membrane and
wastewater recirculation were required as the air-water interface separated anode from cathode.
A polarity inversion aiming at overcoming the pH gradient and enable the biofilm to catalyze
both the anodic and cathodic reaction could be obtained by merely turning the disc array a half
turn.
An electron flow from the submersed half disc to the air exposed half disc established
with the moisture film on the air exposed cathode allowing the ionic charge transfer. As with
other MFC any measured current could be documented to be linked to COD oxidation. The
COD removal caused by the action of the intermittently turning discs was increased by about
30% by merely allowing an electron flow between the anodic and cathodic disc halves. This
result suggests that the treatment performance of traditional RBC may be significantly increased
by using suitable conductive material as the discs.
By raising the cathodic potential from about -500 to -1200 mV (vs. Ag/AgCl) using a
potentiostat the cathodic limitation could be alleviated allowing an increase in electron flow and
COD removal rate to 1.32 kg COD m-3
day-1
(hydraulic retention time 5h). While the COD
removal rate was comparable to that of an activated sludge system, the potentiostatically
supported RBEC removed COD more energy-efficiently than activated sludge systems (0.47 vs.
0.7-2.0 kWh kgCOD-1
), even though it was not optimized. The RBEC could also enable
electrochemically-driven hydrogen gas or methane gas production when operated as a MEC
under fully anaerobic condition.
iii
Overall, this thesis has extended our understanding on how electrochemically active
microorganisms behave in BES. Especially, the discovery of the bidirectional microbial electron
transfer property may shed light not only on BES development, but also on the context of
fundamental microbiology. The new RBEC configuration may widen the functionality or
suitability of BES for a large-scale wastewater treatment application. Nevertheless, further
process optimization is needed. In particular, the cathodic reaction still remains as the key
process bottleneck. Future efforts should thus be oriented towards improving and elucidating in
details the mechanisms of the microbe-cathode electron transfers.
iv
Declaration
I hereby declare that this submission is my own work and that, to the best of my
knowledge, it contains no material previously published or written by another person nor
material which to a substantial extent has been accepted for the award of any other degree or
diploma of the university or other institute of higher learning, except where due
acknowledgment has been made in the text.
Ka Yu Cheng
-----------------------------------------------------
(Signature)
Items derived from this thesis:
1. Patent
―Wastewater Treatment Process‖. Australian Provisional Patent (filed in July 2009). Application number
2009903544. (Chapter 7 and 8)
2. Publications
K. Y. Cheng, R. Cord-Ruwisch and G. Ho. (2007) Limitations of bio-hydrogen production by anaerobic
fermentation process: an overview. American Institute of Physics Conf. Proc. Vol. 941, 264-269. (Cited in
Chapter 1)
K. Y. Cheng, G. Ho and R. Cord-Ruwisch (2008) Affinity of microbial fuel cell biofilm for the anodic potential.
Environmental Science and Technology. Vol. 42(10), 3828-3834. (Chapter 3)
K. Y. Cheng, R. Cord-Ruwisch and G. Ho. (2009) A new approach for in situ cyclic voltammetry of a
microbial fuel cell biofilm without using a potentiostat. Bioelectrochemistry. Vol. 74, 227-231 (Chapter 4)
K. Y. Cheng, G. Ho and R. Cord-Ruwisch. (2010) Anodophilic biofilm catalyzes cathodic oxygen reduction.
Environmental Science and Technology. Vol. 44(1), 518-525. (Chapter 6)
K. Y. Cheng, G. Ho and R. Cord-Ruwisch. Rotatable bio-electrochemical contactor (RBEC)-A novel
wastewater treatment hybrid technology of rotating biological contactor and bioelectrochemical system.
To be Submitted. (Chapter 7)
K. Y. Cheng, G. Ho and R. Cord-Ruwisch. A rotatable bio-electrochemical contactor (RBEC) enables
electrochemically driven methanogenesis. To be Submitted. (Chapter 8)
3. Award
HUBER Technology Prize 2008 (first prize). In IFAT 2008 - 15th International Trade Fair for Water - Sewage
- Refuse - Recycling, Munich, Germany. (Part of Chapter 2)
v
Acknowledgments
I wish to express my gratitude to a number of individuals. Without their support and help
over the past three years, my PhD journey would not be as meaningful and unforgettable as it is.
First of all, my thanks should go to my PhD supervisors who have given me the greatest
freedom ever in steering the research direction throughout the thesis. I would like to express my
deepest gratitude to each of them:
o Prof. Goen Ho — who has constantly offered me opportunities, supports, guidance,
advices and encouragement… Thanks Goen!
o Dr. Ralf Cord-Ruwisch — who has changed my view of dealing with sciences and has
offered me his unlimited inspirations on both sciences and philosophies…,. ―Suck &
See!‖ I am also thankful for the opportunities to participate with his various research
and teaching activities, also for his encouragement, coffee and fun times … Thanks Ralf!
Now, I would like to express my immense gratitude to:
The examiners of this thesis for their valuable expert comments:
o Prof. Lars Angenent, Department of Biological and Environmental Engineering, Cornell
University, USA;
o Prof. Sang-Eun Oh, Department of Biological Environment, Kangwon National
University, South Korea; and
o Prof. Uwe Schroder, Sustainable Chemistry & Energy Research, Institute of Ecological
Chemistry, Technical University-Braunschweig, Germany;
Murdoch University for offering me the full scholarship and living allowance. Thanks
Goen and Ralf again for topping up my scholarship. Without these financial supports I
would not be able to thrive through my research journey in such a beautiful and peaceful
country.
Commericialization office of Murdoch Univeristy for funding (Discovers Grant) the
works in Chapter 7 and 8 of this thesis. Thanks Dr. Patty Washer and Ms. Sam Dymond
of this office for their expert supports on the patent application in the later phase of the
thesis.
My office and laboratory colleagues for their encouragement and precious friendship —
Isabella, Davina, Nora, Lee Walker, Wipa, Donny, Mitch, Suwat, Ying, Raj, Liang and
Chia …Thanks Mates!
All technical staff from the university mechanical workshop. Special thank is given to
Mr. Fritz Wagen for his brilliant contribution to the design and manufacture of the RBEC
reactors. Also, thanks Fritz for showing me your amazing homemade little Jet-engine in
your own backyard workshop!
vi
Mr. John Snowball from the electronic workshop for his excellent assistance on the
manufacture of various electronic devices throughout the project. His valuable technical
advice is also highly appreciated.
Dr. Lucy Skillman for her technical assistance with the biofilm analysis and Mr. Gordon
for his technical assistance on using the scanning electron microscope.
Dr. Korneel Rabaey from the Advanced Water Management Centre, The University of
Queensland for his kind assistance in manufacturing the high quality bioelectrochemical
reactor (and the accessories such as the graphite granules and the cation exchange
membrane) used in this thesis and also for his valuable advice on the topic.
Staff in the School of Environmental Science and the School of Biotechnology and
Biological Sciences for creating such a friendly, dynamic and harmonic environment to
work in and also for allowing me to use various facilities and equipments.
Prof. Jonathan Wong (Hong Kong Baptist University), my honors and masters supervisor
who has constantly given me encouragement and advice since the first day of my
research journey.
Last but not the least; my deepest appreciation are given to my wonderful family: my parents,
my brothers (Karson and Kelvin), my wife Suki and her family. – Thanks for your
understanding, encouragement in every part, continuous support & love…
Finally I wish to share the joy of completing this work with all of the aforementioned
individuals, without their ‗source of sustainable energy‘, completing the thesis alone would be
highly ―unfeasible‖!
Thank You!
Ka Yu
Bateman - Perth
Winter 2009
vii
Table of Contents
Abstract i Declaration iii Acknowledgements v Table of Contents vii
Page
Chapter 1 Introduction and Aim 1 1.1 Background 2
1.1.1 We need to reduce Global Energy Consumption 2
1.1.2 Wastewater Treatment: From "Energy-to-Waste" to "Energy-from-Waste" 3
1.1.2.1 Activated Sludge Processes: Energy-to-Wastewater 4
1.1.2.1.1 Bringing Oxygen from Air to Wastewater requires Energy 4
1.1.2.1.2 Electron Transfer instead of Mass Transfer 4
1.1.2.2 Options for Wastewater-to-Energy 6
1.1.2.2.1 Methanogenic Anaerobic Digestion 6
1.1.2.2.2 Fermentative Hydrogen Production 7
1.1.3 Bioelectrochemical Systems for Sustainable Wastewater-to-Energy 9
1.1.4 Overview of Bioelectrochemical Systems 10
1.1.4.1 BES: A Fast Growing Field of Research 10
1.1.4.2 Basic Features of BES 10
1.1.4.3 Principles and Thermodynamics of Bioelectrochemical Conversion in BESs 12
1.1.4.3.1 MFC and MEC: Similarities and Differences 14
1.1.4.3.1.1 Similarities: Anodic Oxidation is catalyzed by Microbes 14
1.1.4.3.1.1.1 Dilemma: A Lower or a Higher Anodic Potential? 15
1.1.4.3.1.2 Difference: Exergonic or Endothermic Cathodic Reduction? 16
1.1.4.4 Microbial Oxidation of Organics Using an Insoluble Electrode as Electron Acceptor
17
1.1.4.4.1 Exocellular Electron Transfer and Mediators 18
1.1.4.4.1.1 Self-Mediated Exocellular Electron Transfer 19
1.1.5 Limitations in BES Processes 20
1.1.5.1 The Use of Membrane Separators leads to pH Splitting Phenomenon 20
1.1.5.1.1 pH Splitting reduces BES Performance 21
1.1.5.1.1.1 A Fundamental Understanding on the Effect of pH Change on BES Performance
21
1.1.5.1.2 Conventional and State-of-the-Art Approaches for pH Control in BES 23
1.1.5.2 Poor Cathodic Oxygen Reduction in MFC 24
1.2 Aim and Scope of the Thesis 25
Chapter 2 Establishing an Anodophilic Biofilm in a MFC with a Ferricyanide-Cathode and Online pH Control
27
2.1 Introduction 28
2.2 Experimental Section 30
2.2.1 Microbial Fuel Cell Construction 30
2.2.2 Bacterial Inoculum and Medium 31
2.2.3 Start-Up and Operation of a Ferricyanide-Cathode MFC 32
2.2.4 Start-up of Anodophilic Activity in a Potentiostatic-Coupled MFC 33 2.2.5 Performance of a Dissolved Oxygen-based Catholyte 34
2.2.5.1 Effect of Catholyte pH and Phosphate Buffer Concentration 34
2.2.6 Calculation and Analysis 36
2.2.6.1 Determination of Voltage, Current and Power Generation 36
2.2.6.2 Polarization Curve Analysis 36
2.2.6.3 Acetate Analysis 36
2.3 Results and Discussion 38
viii
2.3.1 Quick Start-up of MFC using Ferricyanide-Cathode and Continuous pH-Static Control
38
2.3.1.1 The Cathodic Oxidation Power could influence the Anodic Microbial Activity only at a Low External Resistance
40
2.3.2 Potentiostatically-Controlled MFC also revealed Quick Anodophilic Activity Onset from the same Activated Sludge Inoculum
41
2.3.2.1 Evolution of Anodic Current after a short Lag-phase of only One Day 42
2.3.2.2 Anodophilic Bacteria: Biofilm instead of Suspended Cells 43
2.3.2.3 The Anodophilic Biofilm Established in the Potentiostatically Controlled MFC also gave similar Power Output as the Ferricyanide-Cathode MFC
43
2.3.2.4 SEM reveals only Thin and Low Biomass Density at the Highly Active Biofilm-Anode
45
2.3.3 Performance of a Dissolved Oxygen-based Cathode 46
2.3.3.1 Acidified Catholyte increases Open Circuit Voltage of the MFC 47
2.3.3.2 Phosphate Buffer improves the Dissolved Oxygen-Cathode Performance 48
2.3.3.3 Ferricyanide-Cathode Outperforms Dissolved Oxygen-Cathode 50
2.4 Conclusion and Implication 51
Chapter 3 Affinity of Microbial Fuel Cell Biofilm for the Anodic Potential 53
3.1 Introduction 54
3.2 Experimental Section 56
3.2.1 Microbial Fuel Cell Start-up and On-line Process Monitoring 56
3.2.2 Calculation and Analysis 58
3.2.2.1 Determination of Voltage, Current and Power Generation 58
3.2.2.2 Measuring the Effect of Anodic Potential on Microbial Activity by varying External Resistance
58
3.3 Results and Discussion 59
3.3.1 Characteristics of the MFC 59
3.3.2 Initial Changes of Resistors result in Steady States of Different Microbial Activities
59
3.3.3 At a Certain Potential Range the Dependency of Microbial Activity on Anodic Potential is Less Defined
60
3.3.4 Apparent Maximum of Microbial Activity at Relatively Low Anodic Potentials 62
3.3.5 An Activity Maximum is only obtained when moving from Low to Higher Anodic Potentials
63
3.3.6 Open Circuit Drop in Anodic Potential 64
3.3.7 Detailed Interpretation of the Dependency of Microbial Activity on Anodic Potential
65
3.3.8 Redox Capacitance as an Explanation of Apparent Current Maximum 68
3.3.9 Implication of Findings on the Design and Operation of MFCs 68
Chapter 4 A New Approach for in-situ Cyclic Voltammetry of a Microbial Fuel Cell Biofilm without using a Potentiostat
69
4.1 Introduction 70
4.2 Experimental Section 72
4.2.1 Underlying Principle of the Proposed Method 72
4.2.2 Microbial Fuel Cell 72
4.2.3 Cyclic Voltammetry by Feedback-Controlling the External Resistance 73
4.2.4 Cyclic Voltammetry using a Three-electrode Potentiostat 76
4.3 Results and Discussion 77
4.3.1 Feedback Controlling the External Resistance of a MFC enables Cyclic Voltammetry of MFC Biofilm
77
4.3.2 Limitations and Implications of the New Method 80
Chapter 5 An Anodophilic Biofilm prefers a Low Electrode Potential for Optimal Anodic Electron Transfer: A Voltammetric Study
82
ix
5.1 Introduction 83
5.2 Experimental Section 85
5.2.1 Anodophilic Biofilm and Growth Medium 85
5.2.2 Construction and Operation of Bio-electrochemical Cell 85
5.3 Results and Discussion 88
5.3.1 Only Actively Metabolizing Biofilm Exhibited Electrochemical Activity in Cyclic Voltammetry
88
5.3.2 Redox Mediators of the Active MFC Biofilm were Located Far Away from the Electrode Surface
90
5.3.3 Detail Interpretation of the Anodophilic Properties of the MFC Biofilm in CV 91
5.3.4 Step-Change of Electrode Potential Signified the Existence of an Optimal Electrode Potential for Current Production from the Biofilm
93
5.3.5 Implication of Findings 95
Chapter 6 Alternating Bio-Catalysis of Anodic and Cathodic Reactions alleviates pH Limitation in a BES
97
6.1 Introduction 98
6.2 Experimental Section 101
6.2.1 Electrochemically Active Biofilm and Growth Medium 101
6.2.2 Construction and Monitoring of the Bioelectrochemical System 101
6.2.3 Operation of the Bioelectrochemical Process 103
6.2.3.1 General Operation 103
6.2.3.2 Alternating Supply of Acetate and Oxygen to the Anodophilic Biofilm 103
6.2.3.3 Cathodic Electron Balances 105
6.2.4 Catalytic Effect of the Anodophilic Biofilm on Cathodic Oxygen Reduction 105
6.3 Results and Discussion 106
6.3.1 Acidification of Electrolyte diminishes Anodic Current Production 106
6.3.2 Alternating Supply of Acetate and Oxygen to a Biofilm limits Anodic Acidification
108
6.3.3 Operating the Described MFC at +200 mV did not enable a Cathodic Electron Flow
110
6.3.4 Cathodic Electron Balance 110
6.3.5 Catalytic Effect of the Anodophilic Biofilm on the Cathodic Reaction 111
6.3.6 Expected Power Production of MFC with Anodophilic Bacteria at both the Anode and Cathode
113
6.3.7 Implication of the Findings 113
6.3.7.1 Potential Benefits of the Proposed Concept to developing Sustainable MFC Processes
113
6.3.7.2 Anodophiles Catalyzing the Cathodic Reaction 114
Chapter 7 A Scalable Bioelectrochemical System for Energy Recovery from Wastewater — Rotatable Bio-Electrochemical Contactor (RBEC)
116
7.1 Introduction 117
7.2 Experimental Section 118
7.2.1 Construction of the RBEC Reactor 118
7.2.2 Design of Two Half Discs Serving as Anode and Cathode 119
7.2.3 Process Monitoring and Control 121
7.2.4 Bacterial Inoculum and Synthetic Wastewater 123
7.2.5 Reactor Operation 123
7.2.5.1 Start-up as a Microbial Fuel Cell 123
7.2.5.2 Coupling the RBEC with a Power Source 124
7.2.6 Scanning Electron Microscopy of Biofilm-Electrode 124
7.3 Results and Discussion 126
7.3.1 Operation as a Microbial Fuel Cell for Electricity Generation 126
7.3.1.1 Increased Anodophilic Activity increases Power Output over Time 126
x
7.3.2 In Situ Supply of Oxidizing Power via Conductive Contactor (as Electron Flow) increases COD Removal
128
7.3.2.1 Parasitic Current decreases Coulombic Recovery 130
7.3.3 Sequential Flipping the Electrode Discs allows Alternate Current Generation
132
7.3.4 Coupling the RBEC with an External Power Source to achieve Higher Current
134
7.3.4.1 Electrochemically Assisted Anode Facilitates the Establishment of Anodophilic Biofilm
134
7.3.4.2 Sequential Flipping the Electrochemically-Assisted Discs establishes Anodophilic Biofilm on both Half Discs
136
7.3.5 Intermittent Flipping of the Discs avoids the Continuous Alkalization of the Cathode
137
7.3.6 The Established Biofilm could catalyze a Cathodic Oxygen Reduction 138
7.3.7 Energy Evaluation of the Electrochemically Assisted RBEC process 140
7.4 Concluding Remarks 140
Chapter 8 A Rotatable Bioelectrochemical Contactor (RBEC) enables Electrochemically Driven Methanogenesis
142
8.1 Introduction 143
8.2 Experimental Section 144
8.2.1 Bacterial Seeding Inoculum and Synthetic wastewater 144
8.2.2 Anoxic Operation of the Reactor Headspace of the Electrochemically-Assisted RBEC
144
8.2.3 Gas Production and Measurements 146
8.2.4 Examination of Electrode Biofilm using Scanning Electron Microscopy 146
8.3 Results and Discussion 148
8.3.1 Anoxic Cathode enables Hydrogen Gas Formation 148
8.3.2 Methane instead of Hydrogen was the Predominant By-product after Prolonged Anoxic Operation
151
8.3.3 Is Hydrogen a Key End-product of the Cathodic Reduction Reaction in the Methane-Producing RBEC?
154
8.3.4 Energy Balance 156
8.4 Implication of Findings 157
Chapter 9 Conclusions and Outlook 158 9.1 Conceptual Progression of the Thesis 159
9.1.1 A Highly Anodophilic Biofilm can be Easily Established from Activated Sludge
159
9.1.1.1 MFC Power Output is Undermined by Limitations other than the Metabolic Capacity of the Bioanode
159
9.1.1.2 Wastewater Treating-BES requires Innovative R&D Approaches 161
9.1.2 Anodophilic Biofilm Affinity for the Anodic Potential 163
9.1.3 The Use of Computer Feedback for Dynamic BES Process Control 163
9.1.4 The Rotatable Bioelectrochemical Contactor: A New Option for Large-Scale Wastewater Treatment
164
9.1.4.1 RBEC as a MEC for Enhanced Cathodic Oxygen Reduction, Hydrogen or Methane Production
167
9.2 Final Remarks 168
References 169
Appendix 183 Author’s Curricular Vitae 198
- 1 -
1 Introduction and Aim
Chapter Summary
Every day, we turn ‗clean‘ water into ‗waste‘ water. For sanitation reasons, wastewaters
need to be treated in order to protect human health and the environment from water borne
diseases, and pollution such as eutrophication. However, traditional wastewater treatments such
as activated sludge processes are energy intensive.
Owing to the ever increasing concerns of global energy crisis and its associated
environmental impact, there is a growing need to develop a more energy-efficient wastewater
treatment process. A more radical goal is to recover useful energy from the treatment process.
Recently, bioelectrochemical systems (BESs): microbial fuel cell (MFC) and microbial
electrolysis cell (MEC) are well perceived as the prime technologies to achieve this goal (Logan
2005a; Lovley 2008; Rabaey and Verstraete 2005; Rozendal et al. 2008a).
In the past few years, significant research progress has already been made towards this
goal. However, as BESs are highly complex systems involving multi-disciplinary knowledge
such as microbiology, electrochemistry, material science and engineering, many hurdles still
need to be overcome before the technology becomes practical. The overall aim of this thesis is to
generate understanding that will be helpful in the development of an energy recovering
wastewater treatment process using BESs. The thesis is broadly divided into two main themes.
Firstly is the study of the fundamental aspect of BESs aiming to explore the potential of BESs
for electricity generation and to identify process bottlenecks. The second theme is to develop
practical solutions and to design a practical reactor configuration to overcome the established
bottlenecks.
This chapter aims at reviewing the current state of the art in research and development
related to energy recovery from wastewater using BESs. The basic principles of BESs are
outlined. Emphasis is given on BES that can convert organics into electrical energy directly (i.e.
MFC). Finally, some of the major obstacles of the technology are highlighted. The scope of the
thesis is presented in Section 1.2.
Chapter 1: Introduction and Thesis Aim
- 2 -
1.1 Background – Literature Review
1.1.1 We need to reduce Global Energy Consumption
Of the 13 terrawatts (1012
watts) energy consumed worldwide, approximately 80% (~10
TW) is derived from fossil fuels (e.g. crude oil and natural gas) (Johansson and Goldemberg
2004). Because their formation requires many millions of years, fossil fuels are considered as
non-renewable energy source. Combustion of fossil fuels leads to a continuous increase of
carbon dioxide level in the atmosphere, causing the well-known environmental effects of ―global
warming‖ and climate change (Chamberlain et al. 1982, pp.255; IPCC 2007; Klass 1993)
(Figure 1.1). These effects resulted in a rising sea level due to both expansion of sea water and
the melting of glaciers (IPCC 2007).
Figure 1.1 Historical and projected trends of atmospheric CO2 concentration and average air
temperature. Adapted from Rittmann (2008); original source: Intergovernmental Panel
on Climate Change, IPCC (2007).
Although in the past decade many countries have proposed various approaches to tackle
the problems, still the total world consumption of energy is recently projected to increase by
50% from 2005 to 2030 (EIA 2008). Energy demand in most developed nations (Organization
250
350
450
550
650
750
850
1800 1900 2000 2100Year
Ca
rbo
n D
ioxid
e
(pp
m v
ol.)
15
16
17
18
19
20
Te
mp
era
ture
( oC)
Atmospheric carbon dioxide
Average temperature
Year 2006
Chapter 1: Introduction and Thesis Aim
- 3 -
for Economic Co-operation and Development (OECD) countries) is predicted to expand at an
average annual rate of 0.7%, while energy consumption in the emerging economies of non-
OECD countries (e.g. China and India) is expected to expand at an even higher annual rate of
2.5% (EIA 2008). Of the expanded energy demand, 80-85% will still be derived from fossil
fuels (IEA 2004).
In contrast to the expanding global energy demand, the global supply of fossil fuels has
already peaked and is now declining. A theoretical study has estimated that crude oil and natural
gas reserves will be depleted by the year of 2070 (Klass 2003). Taking also the rapidly growing
world population and the increase of prosperity in developing nations (especially China and
India) into account, the global consumption and demand of energy will increase dramatically
before depletion of fossil fuels (Boeriu et al. 2005).
Clearly, the supply of fossil fuels is unlikely matched with our ever increasing energy
demand. The consequences will not only be an increased energy price leading to instability of
the global economy, but more seriously is the irreversible damage of our natural environment.
Hence, reducing fossil energy consumption and developing renewable alternatives are the prime
objectives of mankind as well as of this PhD thesis.
1.1.2 Wastewater Treatment: From ―Energy-to-Waste‖ to ―Energy-from-
Waste‖?
Wastewater treatment is of utmost importance to safeguard both human and ecological
health. In most developed and urbanized societies, approximately 1.5% of the total electrical
energy consumption is directed to their municipal wastewater treatment processes (Logan 2008).
Therefore, developing a more energy-efficient wastewater treatment can alleviate such energy
demand. In fact, most of the wastewaters (e.g. human, animal, and industrial) already contain a
large amount of chemical energy stored in the form of dissolved organic matter (Angenent et al.
2004). This energy not only can be harnessed for ―cleaning up‖ the wastewater itself, but it can
also be converted into other useful energy forms, available for human consumption again. As
dissolved organic matters in wastewaters are primarily derived from solar energy (via
photosynthesis), wastewaters can be seen as a potential renewable energy resource.
Chapter 1: Introduction and Thesis Aim
- 4 -
1.1.2.1 Activated Sludge Processes: Energy-to-Wastewater
Activated sludge (AS) process is currently the most widely adopted biological
(secondary) process for municipal wastewater treatment. However, it is an energy intensive
process in which air (oxygen) is actively transported (either by surface turbines or submerged
diffusers) from the gas-phase (atmosphere) into the liquid-phase (wastewater) in order to
maintain the wastewater aerobic and the activated sludge (a highly concentrated bacterial
consortium) in suspension. The activated sludge degrades the organic contaminants in the
wastewater by using dissolved oxygen as their terminal electron acceptor. For each m3 of
wastewater approximately 8 m3 of air is required (Lee and Lin 2000). Such active aeration
process consumes tremendous amount of energy (about 1 kWh per kg COD removed or about
0.7 kWh m-3
), which commonly leads to about 60-80% of the total operational cost of the entire
wastewater treatment process. Therefore, in order to reduce treatment costs many studies have
been investigating ways to optimize aeration efficiency or to minimize aeration without
compromising treatment efficiency (Chachuat et al. 2005; Chambers et al. 1998).
1.1.2.1.1 Bringing Oxygen from Air to Wastewater requires Energy
The energy demand of aeration in AS processes lies fundamentally on the mass transfer
of air to water. A certain amount of mechanical energy must be supplied to transport a unit
volume of air into a unit mass of wastewater. This energy is directly proportional to the height of
the aeration basin, and hence the depth of the wastewater (Figure 1.2). However, due to limited
oxygen transfer efficiency (OTE) only part of the oxygen in the transferred air is accessible to
the bacteria in the wastewater. Although OTE can be maximized by reducing the size of the air
bubble (Table 1.1), it appears that at least 70% (assuming a maximal OTE of 30% according to
Table 1.1) of the oxidizing power (as oxygen) is inevitably lost in the process. Such inevitable
loss represents a waste of energy in the treatment process.
1.1.2.1.2 Electron Transfer instead of Mass Transfer
Instead of transferring oxygen from a gas phase into a liquid-phase where the bacteria
and electron donors (i.e. organics) are located, ―oxygen-equivalents‖ required for the organic
removal in a wastewater can be indirectly supplied in the form of an electron-accepting (i.e.
oxidizing) surface such as an inert electric conductive electrode in BES processes.
Chapter 1: Introduction and Thesis Aim
- 5 -
Figure 1.2 Aeration energy is directly proportional to water depth. Note: assumes standard
oxygen transfer rate 60 kg O2 h-1
, air blower efficiency 80%, wastewater basin volume
500 m3, air flow rate 1 m
3s
-1, chemical oxygen demand (COD) of wastewater 320 mg L
-1.
Table 1.1 Standard oxygen transfer efficiency (OTE) and oxygenation efficiency (OE) in
activated sludge processes
Bubble Size Bubble Diameter Standard OTE OE
mm % kg O2/kWh
Fine <3 10 to 30 1.2 to 2
Medium 3 to 6 6 to 15 1 to 1.6
Coarse >6 4 to 8 0.6 to 1.2
Adapted from (von Sperling and de Lemos Chernicharo 2005, page 479)
Since electrons are virtually weightless, the transfer of electrons via an electrode does not
incur any mass transfer energy as in AS aeration processes. Further, the fact that oxygen is not
required for the direct oxidation of organics as well as the possibility of direct recovery of
electrical energy from a wastewater treatment process are features that may increase the overall
energy efficiency of a wastewater treatment process. Therefore, BESs are emerging as an
innovative option for wastewater-to-energy in the future.
0.0
0.5
1.0
1.5
2.0
0 2 4 6 8 10
Ae
ratio
n E
ne
rgy
(kW
h· k
g C
OD
-1)
Water Depth (m)
Chapter 1: Introduction and Thesis Aim
- 6 -
1.1.2.2 Options for Wastewater-to-Energy
Prior to BESs, two other biological processes have been proposed for the conversion of
chemical energy in a wastewater to other useful forms of chemical energy (e.g. biogas (methane
+ CO2), hydrogen gas); (1) methanogenic anaerobic digestion and (2) biological fermentative
hydrogen gas production (also known as dark hydrogen fermentation). The by-products of these
two bioprocesses (i.e. biogas/ hydrogen) are eventually converted into electrical energy using
heat combustion engines or hydrogen fuel cells, respectively.
1.1.2.2.1 Methanogenic Anaerobic Digestion
Methanogenic anaerobic digestion is an established technology for treating a variety of
organic wastes including domestic wastewaters (de Mes et al. 2003). Currently, a total of 2,266
registered full scale high-rate methanogenic digesters are currently operating worldwide (van
Lier et al. 2008). Compared to aerobic wastewater treatment, anaerobic treatments generally
consume less energy. Depending on the need for pumping and recycling effluent, the energy
requirements at ambient temperature are in the range 0.05-0.1 kWh m-3
(0.18-0.36 MJ m-3
) (~0.7
kWh m-3
for AS process) (de Mes et al. 2003).
In fact, the generation of biogas enables net energy recovery from the process. For each
kg COD removed about 13.5 MJ CH4 energy is produced, giving 1.5 kWh high-grade electrical
energy (assuming 40% electric conversion efficiency) (van Lier et al. 2008).
Apart from the possibility of recovering useful energy (as biogas) from a wastewater
instead of investing energy in the process, anaerobic wastewater treatment has an additional
advantage over aerobic wastewater treatment — low sludge production. About 5% of the COD
is converted into sludge in anaerobic processes compared to about 30-60% in aerobic processes
(Figure 1.3). Less sludge handling can reduce the overall treatment cost. Further, the residual
anaerobic sludge even has a positive market value as they can be used as the seeding material to
start up another methanogenic treatment process (van Lier et al. 2008).
Nevertheless, the effluents from anaerobic digester typically still contain a high level of
organics (ranging from 0.5 to a few grams of residual volatile fatty acids) (van Lier et al. 2001).
These dead-end organic compounds represent a pool of unrecovered chemical energy in
anaerobic digestion processes. BESs are perceived as the suitable processes to deal with such
Chapter 1: Introduction and Thesis Aim
- 7 -
low organic waste stream (Clauwaert and Verstraete 2008; Rozendal et al. 2008a), and hence,
rather than competing with the existing anaerobic digestion processes, BESs can be seen as a
complementary technology to anaerobic digestion (Pham et al. 2006).
Figure 1.3 Carbon and energy balance in aerobic and anaerobic wastewater treatments.
(Adapted from van Lier et al. (2008))
1.1.2.2.2 Fermentative Hydrogen Production
Fermentative hydrogen production (also known as dark fermentation), on the other hand,
is an emerging waste-to-energy process mainly targeted towards the concept of ―hydrogen
economy‖, a hypothetical future economy in which the primary form of stored energy is
hydrogen (Momirlan and Veziroglu 2005; Rifkin 2002). The underlying principle of
fermentative hydrogen production is very much alike methanogenic anaerobic processes except
that the former process requires inhibition of hydrogen-consuming microorganisms
(hydrogenotrophic methanogens). However, due to thermodynamic constraint of the involved
biochemical pathways, only 10 to 20% of the stoichiometric maximal hydrogen yield (i.e.
maximal 12 moles of H2 per mole of hexose) could be achieved in practice. I have carried out a
review on this topic which has been published (Cheng et al. 2007, see Appendix 1). A more
Aerobic Wastewater
Treatment
Aeration Energy
100 kWh
Influent
100 kg COD
Heat Loss
Sludge
30-60 kg COD
Effluent
10-12 kg COD
Anaerobic Wastewater
Treatment
Influent
100 kg COD
Biogas 40-45 m3 (~70% CH4)
Sludge
5 kg COD
Effluent
10-20 kg COD
Chapter 1: Introduction and Thesis Aim
- 8 -
comprehensive review of fermentative hydrogen production as a wastewater-to-energy method
can also be found in other reviews (Adams and Stiefel 1998; Benemann 1996; Hallenbeck and
Benemann 2002; Hallenbeck and Ghosh 2009).
Although there is a significant industrial demand for hydrogen, producing hydrogen from
wastewater treatment via fermentation with an aim to recover renewable energy is not
appropriate. Compared to methane, hydrogen has a much lower density and boiling point, only
0.09 kg Nm-3
and -252.9 °C, respectively (vs. 0.717 kg Nm-3
and -182.5 °C, respectively for
methane). These physical properties make hydrogen gas less suitable for transport and storage
compared to methane.
It appears that the popular equation ―H2 + O2 → electricity + H2O‖ has somewhat
oversimplified the ―hydrogen economy‖ concept (Bossel 2006). In reality, the equation should
be modified as ―[H2]initial
+ energies… → [H2]final
+ O2 → electricity + H2O‖. From a point
where hydrogen gas is produced (i.e. [H2]initial
) to a point it is converted into electricity (i.e.
[H2]final
), many energy-consuming intermediate steps are involved. These extra energies can only
be obtained from other energy sources (either fossil fuels or other renewable), but certainly not
from the primary hydrogen production step. In other words, a biologically produced hydrogen-
rich biogas requires energy intensive post-treatment such as purification and pressurization
(from 1 bar to about 600 bars) before it is finally delivered to a fuel cell for electricity generation.
How do we justify the amount of CO2 released if these extra energies have to be derived from
nonrenewable fossil fuels?
Hydrogen generation using bioelectrochemical system (BES) has been proposed in the
literature since 2005 (Liu et al. 2005; Rozendal et al. 2006b). In these pioneering studies, (non-
fermentable) acetate was successfully used for the generation of hydrogen leading to an overall
increase in hydrogen yield. BES is therefore considered as a very promising technology for
turning wastewaters into a useful form of energy (apart from hydrogen, even electricity can be
directly recovered from wastewater using BES), and it will be introduced in the remaining
sections of this chapter.
Chapter 1: Introduction and Thesis Aim
- 9 -
1.1.3 Bioelectrochemical Systems for Sustainable Wastewater-to-Energy
In nature, energetic of biological systems involves electron flow. Electrons flow from an
electron donor that becomes oxidized, to an electron acceptor that becomes reduced. Hence, it is
possible to manipulate biological processes involved in a wastewater treatment process using
electrochemical methods.
The early discoveries of a bacterial capacity to oxidize dissolved organic matters using
an electron accepting electrode (Davis and Yarbrough 1962; Potter 1911) had provoked the
recent development of bioelectrochemical systems (BESs), including microbial fuel cell (MFC)
and microbial electrolysis cell (MEC) processes. BESs allow a direct conversion of chemical
energy stored in a chemical compound into a useful form of energy, with MFC used for a net
production of electricity and MEC for a production of hydrogen gas as a by-product (Rozendal
et al. 2008a).
Recovering energy from wastewaters using BESs is promising because of their high
energy conversion efficiencies compared to all other energy conversion bioprocesses (Rabaey et
al. 2005c). Further, BESs can treat wastewaters that are not suitable for anaerobic digestion
processes (e.g. low-strength wastewater, wastewater composed with mainly volatile fatty acids,
etc) (Rittmann 2008; Watanabe 2008).
Similar to anaerobic digestion, BES also enables removal of organics (anodic oxidation)
with a much lower sludge generation yield (ranging from 2.4 to 26.5 times lower) compared to
aerobic activated sludge processes (Clauwaert et al. 2008a). For instance, a yield of 0.02-0.05 g
biomass-C (g-substrate-C)-1
with acetate as the electron donor substrate (Aelterman et al. 2008);
0.07-0.22 g biomass-COD (g substrate-COD)-1
with glucose as the electron donor substrate
(Rabaey et al. 2003); while 0.53 g biomass-COD (g substrate-COD)-1
for activated sludge
wastewater treatment processes (Verstraete and van Vaerenbergh 1986). Hence, BESs are
emerging as a clean technology for energy recovery from wastewaters.
Chapter 1: Introduction and Thesis Aim
- 10 -
1.1.4 Overview of Bioelectrochemical Systems
1.1.4.1 BES: A Fast Growing Field of Research
In just a few years course of the author‘s PhD study (2006 - 2009), BES research has
been rapidly expanding in the scientific community worldwide. This results in a significant
increase in both the number of published reports and our current knowledge of BES, making it a
challenging task to prepare an up-to-date literature review on the topic. Hence, readers are
kindly referred to the current literature for a more comprehensive overview of specific topic of
BES. In particular, several excellent and extensively cited articles are recommended: (Clauwaert
et al. 2008a; He and Angenent 2006; Logan et al. 2008; Logan et al. 2006; Lovley 2006b; Pham
et al. 2006; Rabaey et al. 2005c; Rabaey and Verstraete 2005; Rozendal et al. 2008a; Schroder
2007).
The following sections review the underlying principles of BES that are prerequisites for
this thesis, and pinpoint the major limitations hindering the up-scaling potential of the BES.
Only BESs using bacteria as the electrochemical ―living catalyst‖ are concerned here.
1.1.4.2 Basic Features of BES
To realize the conversion of chemical energy from wastewater into electrical energy
(electron flow), all BES should have several basic features (Figure 1.4). Firstly, an electron
donor (e.g. organics present in a wastewater) is oxidized by an electrochemically active
anodophilic biofilm which can subsequently transfer the liberated electrons to an electrode (here
anode). Secondly, the electrons at the anode are driven by a potential gradient toward a counter
electrode (here cathode) via an external conductive circuit. In MFC, such electron flow
generates a net electrical energy if a suitable resistive load such as a resistor is located in the
circuit. The MFC cathode can spontaneously reduce an oxidized species (e.g. oxygen). While in
MEC, an external electrical energy from a power source (e.g. DC supplies or potentiostats) is
required to trigger a cathodic reduction (e.g. reduction of protons for H2 formation).
In BES the cathodic reduction can be of either abiotic or biotic nature, where a suitable
bacterial culture (usually exists as a biofilm) catalyzes the electron flow from the cathode to a
terminal electron acceptor (Clauwaert et al. 2007b; Rabaey et al. 2008; Rozendal et al. 2008b).
To date, the so-called biocathode has become a preferred option for BES operation as the need
Chapter 1: Introduction and Thesis Aim
- 11 -
of using expensive and non-sustainable metal-based chemical catalysts (e.g. platinum) can thus
be excluded (He and Angenent 2006; Rabaey et al. 2007).
Figure 1.4 Schematic drawing of the basic features of microbial fuel cell (MFC) and microbial
electrolysis cell (MEC). Red.= reduced bacterial substrate (electron donor); Ox.=
oxidized electron acceptor.
Finally, in order to sustain electron flow (flow of negative charge) from the anode to
cathode in the BES the ionic charges between the two half cells must be balanced (neutral). In
other words, for each electron flowing across the external circuit, either a positively or a
negatively charged ionic species must be transferred from the anode to the cathode or vice versa
(Figure 1.4).
It is worth to point out that in most reported BES studies so far, ion exchange membranes
(e.g. cation exchange membranes, CEMs; anion exchange membranes, AEMs; bipolar
membranes, BPMs) are used to physically separate the reductants (at anode) and oxidants (at
Direction of electron flow
Cation
Red. Ox.Ano-
Bact.
Cath-
Bact.
(Optional)
Anode
Anionor
Cathode
e- e-
e- e-
e- e-
Ion exchange membrane(Optional)
MFC: Power User
(e.g. Resistor)
MEC: Power Supply
Electrolyte Electrolyte
Chapter 1: Introduction and Thesis Aim
- 12 -
cathode) and to selectively allow ion mobilization between the two half cells. However, it can
also be excluded. Some recent studies have proven that both MFC and MEC could operate
without an ion exchange membrane (Call and Logan 2008; Clauwaert and Verstraete 2008; Hu
et al. 2008; You et al. 2007).
Indeed, apart from a high capital cost the use of membranes has several drawbacks on
BES‘s performance. The most well known limitations are the increase in ohmic resistance and
the establishment of a membrane pH gradient (Harnisch et al. 2008; Rozendal et al. 2006a). This
issue will be further discussed in Section 1.1.5.1.
Now, we can move on to the fundamental principle of how and why electrons can flow
in BES.
1.1.4.3 Principles and Thermodynamics of Bioelectrochemical Conversion
in BESs
The likelihood of any chemical reaction to occur in an electrochemical system under
standard or actual conditions can be evaluated by calculating the Gibb‘s free energy change (ΔG)
of the reaction (Bard et al. 1985; Chen 2003). Similarly, this principle can also be applied in
BESs. Electrical energy in BES is spontaneously generated only when the overall reaction is
thermodynamically (or energetically) favorable (ΔG < 0). Yet, it may be more convenient to
evaluate the reaction in terms of the overall cell electromotive force, Eemf (V), which is defined
as the potential difference between cathode (i.e. positive terminal) and anode (i.e. negative
terminal) of the BES:
Eemf (V) = E cathodic (V) – E anodic (V) (eq. 1-1)
Where, E cathode = emf of a specific reaction taking place at cathode
E anode = emf of a specific reaction taking place at anode
The maximum work (Wmax) that a system can perform is related to the Gibbs free energy
changes (∆G). This can be estimated from Eemf according to the following relationship (Logan et
al. 2006):
Chapter 1: Introduction and Thesis Aim
- 13 -
Wmax = Eemf·Q = Eemf·n·F= -∆G (eq. 1-2)
Where, Wmax = maximum theoretical work
Eemf = potential difference between the anode and cathode
Q = charge transferred in the reaction, expressed in Coulomb (C)
n = number of electrons per reaction
F = Faraday‘s constant (96,485 C mol-1
)
Rearranging equation 1-2 gives:
Eemf = − ∆G
nF (eq. 1-3)
Under standard condition (i.e. 298K, 1 atm pressure, 1 M concentration for all species) gives:
Eemf𝑜 = −
∆G0
nF (eq. 1-4)
Where E0
emf = standard cell electromotive force. Based on the above equations, the overall
electromotive force of a particular reaction under any specific condition can be calculated
according to the Nernst equation (eq. 1-5):
Eemf = Eemf𝑜 −
RT
nFln(П) (eq. 1-5)
Where, R = universal gas constant (8.314 x 10-3
kJ∙mol-1
∙K-1
);
T = absolute temperature (K);
П = reaction quotient calculated as the activities (i.e. concentrations at dilute
concentrations) of the products divided by those of the reactants.
For a reaction of: aA + bB → cC + dD, the reaction quotient (П) can be defined as the
following (Bard and Faulkner 2001):
П =[𝐶]𝑐 [𝐷]𝑑
[𝐴]𝑎 [𝐵]𝑏 (eq. 1-6)
Chapter 1: Introduction and Thesis Aim
- 14 -
1.1.4.3.1 MFC and MEC: Similarities and Differences
1.1.4.3.1.1 Similarities: Anodic Oxidation is catalyzed by Microbes
The first step of the microbial transformation of chemical energy into electrical energy in
both MFC and MEC is similar (Figure 1.5). In which, a reduced (more negative redox potential)
electron donor substrate is taken up by the bacteria located at close proximity (or directly
attached) to an anode, which has a higher (more positive) redox potential compared to the
bacteria itself. The bacteria will then catabolize the electron donor via a series of biochemical
redox reactions inside the cell (e.g. TCA cycle). In these redox processes, the bacteria must
conserve a certain amount of energy (i.e. Gibbs free energy) to meet their cellular needs (e.g. for
cell maintenance and growth) (Thauer et al. 1977).
Figure 1.5 A schematic diagram showing the thermodynamics of microbial fuel cell (MFC) and
microbial electrolysis cell (MEC). Note: This representation neglects electrochemical
losses (also known as overpotential) based on ohmic resistances (electrode and
electrolyte), concentration (mass transfer limitation) and activation polarizations and
kinetic constraints.)
Redox Potential (V)
+ve
e- Donor
(Acetate)
Microbe
Anode
-0.28
+0.82
ΔGo’Elec= -ve
-0.42 e- Acceptor
(H+/ H2)
ΔGo’bacteria = -ve
e- Acceptor
(O2/ H20)
Energy
Input
Energy
Output
ΔGo’Elec= +ve
-ve
MEC
MFC
Load(e.g. resistor)
Power supply
Cathode
ΔGo'Max. Output = -ve
ΔGo'Min. Input = +ve
1.1 V
0.14 V
Anion
Cation
or
Chapter 1: Introduction and Thesis Aim
- 15 -
1.1.4.3.1.1.1 Dilemma: A Lower or a Higher Anodic Potential?
From our energy recovery standpoint, such bacterial energy gain (denoted as ΔGo‘
bacteria
in Figure 1.5) can be considered as an ―energy loss‖ because it cannot be transformed into
electrical energy in the BES (Schroder 2007). Thus, a dilemma exists when determining which
processes (bacterial energy gain? or an anode with a lower anodic potential?) should be favored
more when operating a BES (Figure 1.6).
Nevertheless, it is clear that in order to maximize electricity generation in MFC or to
minimize the extra energy input in MEC (both are denoted as ΔGo‘
Elec in Figure 1.5), a lower
amount of ΔGo‘
bacteria is preferred as it keeps the anode at a lower redox state and hence the
anode can be more reducing. Further, if the bacterial energy gain becomes too large (at ―too high‖
anodic potential), the electrical energy output may become diminished and electrons from the
substrate may be diverted toward an assimilation metabolism, leading to excessive biomass yield
but reduced electron recovery. Therefore, if the anodic bacteria could still benefit from a small
potential gradient (between the electron donor substrate and the anode) a lower operating anodic
potential is preferred.
Figure 1.6 Effects of anodic potential on the theoretical cell voltage and bacterial energy gain
(Gibbs free energy change, ΔGo‘
). Notes: ΔGo‘
data refer to standard condition, pH 7 and
298K; cell voltage (V) = cathodic potential – anodic potential.
-1.0
-0.5
0.0
0.5
1.0
1.5
-600 -400 -200 0 200 400 600
Anodic Potential (mV vs. Ag/AgCl)
ΔG
o'(k
J m
ol -1)
-900
-800
-700
-600
-500
-400
-300
-200
-100
0
100
Ce
ll V
olta
ge
(V
)
Cell Voltage: Coupled with an O2/H20 Cathode (+0.62 V |Ag/AgCl)
Gibbs free energy change
Cell Voltage: Coupled with a Poor Cathode (0 V |Ag/AgCl)
Ba
cte
rial E
ne
rgy G
ain
BE
S E
ne
rgy O
utp
ut
Chapter 1: Introduction and Thesis Aim
- 16 -
Apparently, the above consideration could lead to important research questions… e.g.
―How low (negative) can an operating anodic potential be maintained and yet still allow a
reasonably high (if not maximal) microbial activity (as current)?‖ ―Can we establish an anodic
microbial consortium that has a suitable metabolic and electron transfer pathway to effectively
utilize a very negative anodic potential?‖
1.1.4.3.1.2 Difference: Exergonic or Endothermic Cathodic Reduction?
As depicted in Figure 1.5, the fundamental difference between MFC and MEC lies in the
thermodynamic nature of their cathodic reaction. In MFC, the cathodic potential established by
the reducing power from the anode is sufficient to drive a cathodic reduction of a terminal
electron acceptor (e.g. oxygen, nitrate or ferricyanide) (reaction 1-2 to 1-4 in Table 1.2). Such a
downhill (exergonic) redox reaction (ΔGo‘
Max. Output < 0) enables a spontaneous energy output (as
electricity) if a suitable electrical load (e.g. resistor) is located in the external circuit of the MFC.
Table 1.2 Thermodynamic evaluation of BES with different cathodic reactions (assuming an
anodic acetate oxidation) under standard condition (pH 7, 298K).
Eo’ (V)
a
BES Eemf (V)
BES ΔGo’
(kJ mol-1
) Reaction
Anodic Oxidation
CH3COO- + H2O → CO2 + 7H
+ + 8e
- -0.28 --- --- (1-1)
Cathodic Reduction
O2 + 4H+ + 4e
- → 2H2O +0.82 +1.10 -849 (1-2)
NO3- + 6H
+ + 5e
- → 0.5N2 + 3H2O +0.74 +1.02 -787 (1-3)
Fe(CN)63-
+ e- → Fe(CN)6
4- +0.36 +0.64 -494 (1-4)
H+ + e
- → H2 -0.42 -0.14 +108 (1-5)
a Values adapted from He and Angenent (2006); Eemf values are calculated according to eq. 1-1, Eemf
= E cathodic – E anodic.
On the other hand, the cathodic reduction in MEC requires a more reducing (i.e. negative)
cathodic potential, which cannot be satisfied with the reducing power of the anode (e.g. reaction
1-5 in Table 1.2). An extra amount of reducing power (here a potential of -0.14 V for H2
formation) is required from an external power source (e.g. potentiostat) to kick off the cathodic
Chapter 1: Introduction and Thesis Aim
- 17 -
reaction (ΔGo‘
Min. input > 0). So far, MEC is mostly established for the cathodic generation of
hydrogen gas (i.e. cathodic reduction of protons).
As electrochemically active bacteria catalyze the release of electrons from an organic
electron donor at a higher energy level (i.e. at low (i.e. more negative) anodic potential), the
amount of energy required for hydrogen production is lowered compared to that in a
conventional hydrogen production using water electrolysis method. Typically, such conventional
process requires 1.8 to 2.0 V to electrolyze water. Hence, using MEC to produce hydrogen has
attracted remarkable attention from the industry (Rozendal 2007).
1.1.4.4 Microbial Oxidation of Organics Using an Insoluble Electrode as
Electron Acceptor
Arguably, the most essential phenomenon in BES systems is the capacity of bacteria to
transfer their electrons exocellularly to an insoluble electrode. This bacterial capacity is essential
as it is the key step of converting checmial energy into an electrical energy in all BES systems.
In nature, exocellular electron transfer can be observed in various geochemical environments.
The most well studied examples are the microbial reduction of insoluble Iron (III) oxide (e.g. by
genera of Geobacter) and Manganese (IV) oxide (e.g. by genera of Shewanella) commonly
occuring in anaerobic habitats such as groundwater aquifers and freshwater or marine sediments
(Lovley et al. 2004; Myers and Nealson 1990; Nealson and Saffarini 1994).
Such electron transfer enables the bacteria to conserve energy (e.g. via TCA cycle, ATP
generation by electron transport phosphorylation) to meet their metabolic need. The maximal
amount of energy that the bacteria can possibly conserve is in theory directly proportional to the
redox potential difference between the electron donor (reductant) and the electron acceptor
(oxidant) (Table 1.3).
Table 1.3 Standard reduction potentialsa of various redox couples that are of biological
interest Oxidation/Reduction Couples (mV vs. SHE) (mV vs. Ag/AgCl)
CO2/ Glucose -430 -630
H+/H2 -420 -620
NAD+/ NADH -320 -520
CO2/ Acetate -280 -480
S0/ HS
- -270 -470
SO42-
/ H2S -220 -420
Pyruvate2-
/ Lactate2-
-185 -385
Chapter 1: Introduction and Thesis Aim
- 18 -
2,6-AQDS/ 2,6-AHQDS -184 -384
Menaquinione oxidized/ reduced -75 -275
Pyocyanin oxidized/ reduced -34 -234
Methylene blue oxidized/ reduced +11 -189
Fumarate2-
/ Succinate2+
+31 -169
Thionine oxidized/ reduced +64 -136
Cytochrome b (Fe3+
)/ Cytochrome b (Fe2+
) +75 -125
Fe(III) EDTA/ Fe(II) EDTA +96 -104
Ubiquinone oxidized/ reduced +113 -87
Cytochrome c (Fe3+
)/ Cytochrome c (Fe2+
) +254 +54
O2/ H2O2 +275 +75
Fe(CN)63-
/ Fe(CN)64-
+430 +230
Fe(III) citrate/ Fe(II) citrate +372 +172
NO3-/ NO2
- +421 +221
NO3-/ N2 +740 +540
O2/H2O +820 +620 aStandard biological potential: pH7, 298K. Data mostly from Lehninger et al. (1993).
1.1.4.4.1 Exocellular Electron Transfer and Mediators
Exocellular electron transfer enables microbial oxidation of electron donors (e.g.
organics) using an insoluble electron acceptor such as an anode in BES. Bacteria capable of
catalyzing an electron transfer from a reduced electron donor to an anode are termed as
anodophilic bacteria (throughout this thesis). Other terminologies are also used in the literature.
For example, electricigen (Logan 2008); anode-respiring bacteria (ARB) (Torres et al. 2007);
electrode reducing microorganisms or electrode reducers (Lovley 2008)).
Although earlier studies have emphasized the important role of artificial redox mediators
(e.g. neutral red (Park et al. 2000; Park and Zeikus 2000); 9,10-anthraquinone-2,6-disulfonic
acid (AQDS), safranine O, methylene blue (Sund et al. 2007)) on exocellular electron transfers
in BES (mainly MFC), recent evidence has confirmed that bacteria (in either pure or mixed
culture studies) could establish their anodophilic capacity without using artificial mediators
(systems commonly known as mediator-less MFC) (Chaudhuri and Lovley 2003; Gil et al. 2003;
Kim et al. 2002; Rabaey et al. 2004).
For example, Kim and coworkers (2004) observed in a confocal scanning laser
microscopy study that the microorganisms in the biofilm attached on the anode of their MFC
were densely stained with florescent, indicating a majority of viable bacteria in the biofilm (Kim
et al. 2004a). Since no electron acceptors were supplied except the anode, the survival of those
microorganisms that were not in direct contact with the electrode implies that the bacteria have
developed alternative means to conserve energy using the anode without using artificial
mediators.
Chapter 1: Introduction and Thesis Aim
- 19 -
In fact, from a wastewater treatment standpoint the use of artificial mediators is
impractical as they need to be regularly replenished and may cause water contamination. Hence,
this thesis only considers self-mediated exocellular electron transfer.
1.1.4.4.1.1 Self-Mediated Exocellular Electron Transfer
In order for the anodophilic bacteria to self-mediate electron transfer to an anode, a
―redox link‖ must be developed between the bacteria and the anode. According to Schroder
(2007), an effective redox species must have several basic criteria:
(i) It must be able to physically interact with the electrode surface;
(ii) It must be electrochemically active and exhibits low oxidation overpotential at
the electrode surface;
(iii) Its standard potential should be as close to the redox potential of the electron
donor substrate of the bacterium as possible, or at least be significantly negative
to its final electron acceptor.
Figure 1.7 Exocellular electron transfer enables microbial oxidation of electron donors (e.g.
organics) using an insoluble electron acceptor such as an anode in BES. (A) redox active
mediator(s) serve as electron shuttle; (B) direct contact between bacteria and insoluble
electron acceptor without using soluble mediator(s).
Fe3+ + e- → Fe2+ or;
Mn4+ + 2e- → Mn2+ or;
Anode (oxidized) → Anode (reduced)
Medox
Medred
Bacterium
(1)
(2)
(3)(4)
(5)E donor
Medox
Medred
(7)
(2)
(3)(4)
(5)
(6)
Bacterium
(1)E donor
(1)
(1)
Inso
lub
le e
-Acce
pto
r
(Fe
3+; M
n4
+o
r An
od
e)
e-
A. ExocellularMediator Dependent
(Endogenous or exogenous):
B. ExocellularMediator Independent
(Mediatorless):
Electron flow in BESs
Examples:
Chapter 1: Introduction and Thesis Aim
- 20 -
Remarks: (1) Bacterium acquires electrons from electron donors (i.e. dissolved organic matters);
(2) Bacterium transfers the electrons to an oxidized mediator;
(3) Reduction of the oxidized mediator;
(4) Transfer of electrons from the reduced mediator to an insoluble electron acceptor;
(5) Regeneration of the oxidized mediator;
(6) Electron transfer from a reduced mediator produced by one bacterium to another
bacterium (Pham et al. 2007);
(7) Bacterium transfers electrons directly to an insoluble electron acceptor.
To our current knowledge, anodophilic bacteria may establish such ―link‖ with one of
the following two mechanisms (Figure 1.7): (1) direct physical contact with the anode using a
―link‖ located at the outer membrane (e.g. c-type cytochromes in Geobacteraceae, (Richter et al.
2009); or electrically conductive pili ―nanowires‖, which allow the bacteria located at a longer
distance away from the anode but still able to directly transfer electrons to it (Reguera et al.
2005)); and (2) self-secretion of a soluble redox active shuttle (mediator) that can mediate
electron transfer through diffusion and/or electrostatic interaction (Rabaey et al. 2007).
1.1.5 Limitations in BES Processes
Ongoing BES research efforts mainly focus on fundamental understanding of how
bacteria interact with an insoluble electrode and developing various configurations for different
application. (Clauwaert et al. 2008a; He and Angenent 2006; Logan and Regan 2006b; Logan et
al. 2008; Logan et al. 2006; Lovley 2006b; Rabaey et al. 2007; Rabaey and Verstraete 2005;
Rismani-Yazdi et al. 2008; Rittmann 2008; Rozendal et al. 2008a; Schroder 2007).
In particular, two fundamental limitations are emphasized: (1) the poor cathodic reaction;
and (2) the build-up of a pH gradient between an anode and a cathode over time in non-pH
buffered or poorly pH buffered systems. This so-called pH splitting phenomenon, if not
alleviated, can almost entirely block the electron flow and operation of the MFC.
1.1.5.1 The Use of Membrane Separators leads to pH Splitting Phenomenon
Ideally, either a proton (H+) or a hydroxide ion (OH
-) serves as the migrating ionic
species across an ion exchange membrane (refer to Figure 1.4). However, it becomes well
accepted in the literature that under operating conditions (i.e. pH close to neutral) prevailing in
Chapter 1: Introduction and Thesis Aim
- 21 -
most MFC systems, and also in MEC where protons are continuously consumed for the
generation of hydrogen gas at a cathode, the concentrations of both H+ and OH
- ions are several
order of magnitudes lower compared to most other ionic species in the electrolyte (Harnisch et al.
2008; Rozendal et al. 2008a). This results in a preferential migration of other ionic species to
maintain electroneutrality of the system, causing acidification and alkalization at the anode and
cathode, respectively. At present, there is no real solution available in the literature to
sustainably overcome such membrane associated pH splitting limitation yet (Harnisch and
Schroder 2009).
1.1.5.1.1 pH Splitting reduces BES Performance
pH splitting phenomena could lead to as high as a 6 pH units difference between the two
electrodes (Clauwaert et al. 2008b), severely decreasing the driving force of the system (i.e.
potential difference between anode and cathode) as each unit of pH gradient represents an
overpotential (i.e. potential loss) of 59 mV according to Nernst equation Consequently, the
current is diminished and hence diminishing the wastewater treatment performance.
1.1.5.1.1.1 A Fundamental Understanding on the Effect of pH Change on BES
Performance
The cathodic oxygen reduction in MFC and the cathodic hydrogen formation in MEC are
both proton consuming reactions, i.e.
O2 + 4H+ + 4e
- → 2H2O
H+ + e
- → H2
To understand the effect of pH on the BES performance, the reduction potential of these
cathodic reactions can be modeled according to Nernst equation (eq. 1-5). Since the H+ and OH
-
are always in equilibrium with each other through the water dissociation (Kw=[H+][OH
-] ≈ 10
-14),
net consumption of proton represents a hydroxide ion production. Hence, it is expected that the
localized pH at the cathode disc would increase over time. Assuming the concentration of other
reacting species (e.g O2 and H2) in the above cathodic reactions stay constant over time, then the
Chapter 1: Introduction and Thesis Aim
- 22 -
reduction potentials are solely dependent on the [OH-] and hence the corresponding Nernst
equation can be described as e.q. 1-7 (Rozendal 2007):
Ecathodet>0 = Ecathode
t=0 − RT
nFln
[OH −]t>0
[OH −]t=0
a (e.q. 1-7)
Where Ecathodet=0 and Ecathode
t>0 are the cathodic potential at time = 0 and time > 0, respectively.
The ―a‖ in e.q. 1-7 is the reaction coordinate of the hydroxyl ions in the cathodic reaction. Since
the formation of OH- is directly proportional to the current of the BES (i.e. 1 mol of OH
- is
produced per 1 mol of electron flow across the external circuit), the reaction coordinate, a, in e.q.
1-7 must be exactly equal to ―n‖. Hence, e.q. 1-7 can be further simplified as follow (e.q. 1-8):
Ecathodet>0 = Ecathode
t=0 − RT
Fln
[OH −]t>0
[OH −]t=0 (e.q. 1-8)
Based on this equation, for each unit of pH increase at the cathode (i.e. a 10-fold increase in OH-
concentration) after a certain period of time (t>0), the cathodic potential is decreased by 0.059 V
(i.e. 59 mV) (Figure 1.8):
Ecathodet>0 = Eca thode
t=0 − RT
Fln
[OH −]t>0
[OH −]t=0 (e.q. 1.9)
=Ecathodet=0 −
8.314 298
96485ln 10
= Ecathodet=0 − 0.059 V
As shown in Figure 1.8, alkalinization of the cathode decreases the cathodic potential. If
we assume the anodic reaction of a BES stay constant, a decrease of the cathodic potential
signifies an overall decrease of the driving force (e.m.f.) of the BES (refer to section 1.1.4.3 e.q.
Chapter 1: Introduction and Thesis Aim
- 23 -
1-1). As a consequence, this would diminish electrical energy output of a MFC, or a higher
energy input is required to bring the cathode to the lower potentials in order to trigger the
cathodic reduction in a MEC.
On the other hand, Figure 1.8 suggests a beneficial effect of pH change on the cathodic
potential when the cathode becomes more acidic instead of alkaline. An increase of +59 mV in
the cathode is obtained if the cathode was acidified by one pH unit. However, from a practical
viewpoint, as both the oxygen reduction and hydrogen formation are H+ consuming and hence
OH- producing reactions, the corresponding cathode would only shift to a more alkaline pH.
Hence, pH correction is needed to avoid cathode alkalinization during BES operation.
Figure 1.8 Change in the cathodic reduction potential versus change in cathodic pH. Values
calculated according to e.q. 1-9. ΔpH = pH at t>0 minus pH at t=0.
1.1.5.1.2 Conventional and State-of-the-Art Approaches for pH control in BES
The most commonly used pH control methods in the literature are the addition of buffer
(e.g. phosphate buffer) and by dosing acid/or base to the electrolyte (Rozendal et al. 2008a).
Although they could effectively address the pH limitation in laboratory studies, they are not
preferred for large scale wastewater treatment processes.
Recently, a complete loop-concept has been proposed and validated as a mitigation
strategy to overcome the aforementioned pH limitation during MFC operation (Freguia et al.
2007c). This is an interesting and could be a practical concept where the anodic effluent serves
as the influent for the cathode compartment in a two-compartment MFC. As the biological
y = -59x
-300
-200
-100
0
100
200
300
-5 -3 -1 1 3 5∆pH
∆Ecathode
(mV)
Alkalinization decreasing
Cathodic E
Acidification increasing
Cathodic E
Chapter 1: Introduction and Thesis Aim
- 24 -
oxidation of the electron donor substrate (here acetate) in their MFC anode is a H+ producing
reaction (see reaction 1-1 in Table 1.2), the anodic effluent becomes acidified (pH drops from
about 7.1 to 5.8 - 6.5). The hydraulic connection between the acidified anolyte and the catholyte
could effectively alleviate the alkalinization at the cathode, avoiding the need for pH control or
catholyte replacement (Freguia et al. 2007c). The same concept has been validated in a separate
report with a different BES configuration (Clauwaert et al. 2009).
However, similar to most other configurations reported in the literature the
aforementioned pH control strategy depend heavily on mechanical pumping or recirculation of
wastewater through the system (Clauwaert et al. 2009; Freguia et al. 2007c). Wastewater needs
to be continuously lifted up from the bottom inlet to the top of the anodic compartment, from
which the anodic effluent overflows and trickles over an air-exposed cathode. As the energy
required for lifting up a unit volume (i.e. a unit weight) of wastewater within the anodic
compartment is directly proportional to the compartment‘s height (see Figure 1.2), more energy
needs to be invested to lift the wastewater in a taller system. Such concern has not been
considered, and yet it may set a limit to the up-scaling potential for this kind of configuration.
1.1.5.2 Poor Cathodic Oxygen Reduction in MFC
Apart from the pH splitting limitation, the high overpotential associated with the cathodic
oxygen reduction using graphite and carbon electrodes often limits the MFC performance.
Although modifying the cathode with a platinum catalyst is a proven way to alleviate such
overpotential limitation, the use of platinum cannot be justified in MFC processes because of
costs and environmental impact in its production (Freguia et al. 2007b; Harnisch and Schroder
2009). Recent development of biocathode (bacterial catalyzed cathodic reduction) is seen as a
more sustainable and practical way to overcome the cathodic oxygen reduction problem
(Clauwaert et al. 2007b; He and Angenent 2006; Rabaey et al. 2008). However, the application
of biocathode still remains largely unexplored.
Chapter 1: Introduction and Thesis Aim
- 25 -
1.2 Aim and Scope of the Thesis
The overall aim of this PhD thesis is to explore the potential of BES for energy recovery
from organics in wastewater. The research program is broadly divided into two main themes:
1) Fundamental understanding of BES for electricity generation with emphasis on the
microbe-anode interactions. (Chapter 2 to 5)
2) Developing a practical reactor configuration to overcome the established bottlenecks.
(Chapter 6 to 8)
In general, acetate was used as a model substrate in this thesis as it is the most commonly
selected dissolved organic compound in many other wastewater treatment studies. Further,
acetate is considered as a dead-end product in dark fermentation because it cannot be further
converted into hydrogen due to thermodynamic and biochemical reasons (Cheng et al. 2007).
As mentioned earlier in this chapter (Chapter 1), one of the fundamental and common
steps in both MFC and MEC is the anodic oxidation of organic electron donors. Therefore,
Chapter 2 aims at establishing an anodophilic biofilm from an activated sludge (a commonly
used MFC mixed culture inoculums) in a two-chamber MFC. Since a non-limiting cathode was
essential to evaluate the anodic process (e.g. establishment of biofilm), ferricyanide was used as
the catholyte to offer a stronger and a more stable cathodic reaction.
The anode is an analogue to the terminal electron acceptor for the electrochemically
active bacteria to oxidize organic substrates in a BES. Hence, the energetics of the microbial
oxidation process largely depends on the potential of the anode. However, in the literature the
dependency of the microbial reaction on the anodic potential has not been investigated in detail.
Hence, Chapter 3 investigates the ―attractiveness‖ of the anode to the microbial rate of charging
it with electrons. While the effect of substrate concentration on the rate of microbially catalyzed
reactions is well understood and described as affinity, such relationship has not been described
for insoluble electron acceptors such as an anode. Such relationship was explored by controlling
the external resistance of the MFC.
Although online process control is used for many bioprocesses (e.g. sequencing batch
reactor for wastewater treatment) it is not established for MFC. One of the main features in this
thesis is the application of computer-feedback for BES process control. As an example to this,
Chapter 4 introduces a novel approach for controlling the biofilm-electrode potential by
Chapter 1: Introduction and Thesis Aim
- 26 -
feedback controlling the external resistance of an operating MFC. In analogy to a conventional
cyclic voltammetry study, this new approach could evaluate the redox behavior of an MFC
biofilm but without using a potentiostat. In this way, the MFC can still operate as a ―fuel cell‖
without being ―interrupted‖ by an external device (i.e. potentiostat) which normally does not
belong to the system. Such method could be useful for diagnosing system performance during
MFC power generation.
Cyclic voltammetry is a commonly used electrochemical method to study the electron
transfer between bacteria and the insoluble electrode in a BES. However, the complexity of a
biofilm-electrode system often limits the information obtained from this method. In Chapter 5,
a series of potentiostatic experiments was conducted to further explore the electron transfer
properties of an active anodophilic biofilm. Rather than using standard voltammetry techniques,
this thesis shows the merit of ultra slow scan rates (e.g. 0.01 mV·s-1
) for the study of
bioelectrocatalysis of an electrochemically active biofilm.
Chapter 6, 7 and 8 are devoted towards a practical and sustainable operation of BES for
energy recovery from wastewaters. The concept of using a single biofilm to alternately catalyze
the anodic oxidation of organics and a cathodic oxygen reduction is investigated in Chapter 6.
Its objectives are twofold: (1) to overcome the generic pH drifting problem inherent to the use of
membranes in BES and (2) to alleviate the oxygen reduction overpotential at a (non-catalyzed)
plain granular graphite electrode using an anodophilic biofilm.
Based on the findings of Chapter 6, a novel membrane-less BES configuration, known as
Rotatable Bio-Electrochemical Contractor (RBEC) was designed, fabricated and evaluated for
its performance as a MFC for electricity generation in Chapter 7; and as a MEC for hydrogen or
methane production in Chapter 8.
Finally, Chapter 9 discusses and summarizes the insights obtained in this thesis. Future
research directions and outlook are addressed.
- 27 -
2 Establishing an Anodophilic Biofilm in a MFC with a
Ferricyanide-Cathode and Online pH Control
Chapter Summary
Maintaining a suitable anodic condition in a bioelectrochemical system such as microbial
fuel cell (MFC) is essential for establishing active anodophilic biofilm from the initial inoculum
at an anode. In particular, a pH neutral operating anolyte and a non-limiting MFC cathode are
essential to such acclimation process.
This initial experimental chapter aims at establishing an anodophilic biofilm from a
mixed culture activated sludge inoculum in a pH-controlled, ferricyanide-cathode driven two-
chamber MFC. The MFC was inoculated with activated sludge (10%, v/v) and operated at fed-
batch mode at 30oC for 30 days. A synthetic wastewater amended with acetate (10 mM) and a
potassium ferricyanide solution (50 mM, 100 mM pH 7 phosphate buffer) were used as the
anolyte and catholyte, respectively. The external resistance of the MFC was predominately
maintained at low level (1-5 Ω) to avoid anodic limitation. Anolyte pH was controlled at 7 by
dosing NaOH (1M).
Regular polarization curve analysis indicated an overtime increase in MFC power output.
The rapid buildup of high power densities of the MFC (415 W·m-3
after five days) suggest that
an activated sludge mixed culture inoculum can readily become highly anodophilic at a MFC
anode, provided that factors such as electrolyte pH, external resistance and cathodic oxidizing
power are not limiting. However, the use of ferricyanide catholyte and the continuous alkaline
dosing for pH control are both non-sustainable and therefore unsuitable options for practical
BES application. More suitable alternatives should be explored for a sustainable MFC operation.
Chapter 2: MFC Start-up and Basic Understandings
- 28 -
2.1 Introduction
Start up of bioelectrochemical systems (BES) for simultaneous electricity production and
wastewater treatment usually involves the use of mixed cultures bacterial inoculums such as
activated sludge or anaerobic digester sludge (Logan 2005b; Rozendal et al. 2008a). Under
suitable acclimation condition, it is expected that the BES is able to select and enrich its own
electrochemically active consortia from the mixed culture over time (Kim et al. 2004a; Liu et al.
2008; Rabaey et al. 2003). During this process, the bacteria are able to establish their own
―redox link‖ with the insoluble anode without depending on artificially added redox mediators
(Jang et al. 2004; Schroder 2007). This ―link‖ allows the electrochemically active (or
anodophilic) bacteria to oxidize their electron donors (organics in a wastewater) and transfer the
electrons to the anode for their own energetic need. Such anodic electron transfer process
depends largely on the bacterial activity and the operating condition inside the anode chamber.
Thus, a suitable anodic condition is essential for the establishment of anodophilic biofilm in a
BES during the acclimation period.
One of the challenges in operating a MFC anode is to overcome the acidification of a
poorly pH-buffered anolyte over time (Gil et al. 2003; Pham et al. 2009). Such anolyte
acidification is caused by the proton-producing anodic oxidation of organics (e.g. acetate) (eq. 2-
1).
CH3COO- + 4 H2O 2 HCO3
- + 9 H
+ + 8 e
- (eq. 2-1)
As the electrons are accepted by the anode, the protons are concomitantly released to the anolyte.
Ideally, these released protons will be migrated from the anolyte to the catholyte via a cation-
exchange membrane to maintain the ionic charge balance in the BES. However, under the
prevailing operating environment in most wastewater treating BESs (low ionic strength, neutral
pH, low pH buffer capacity and ambient operating temperature), the proton flux is highly
inefficient. Other cations (e.g. sodium, magnesium or calcium ion, etc) will preferentially
migrate from the anolyte to the cathode (Rozendal et al. 2006a). As a consequent, the anolyte
becomes acidified. It is well accepted in the literature that acidification of the anolyte may
hamper both the electrochemical and biological activity of the anodic process (Clauwaert et al.
2008a; Harnisch and Schroder 2009; Picioreanu et al. 2009). Hence, maintaining a stable and
neutral anolyte pH is expected to allow stable anodic performance.
Chapter 2: MFC Start-up and Basic Understandings
- 29 -
From a bacterial standpoint, the functions of an anode in a BES are twofold: (1) it can
serve as a physical surface for biofilm-forming bacteria to attach and grow; (2) it serves as the
terminal electron acceptor for the electrochemically-active bacteria to derive metabolic energy.
Although it is quite common that electrochemically active bacteria in BES tend to exist as a
biofilm attaching at the anode surface, this is not an essential prerequisite for anodic electron
transfers because suspended bacteria could also generate anodic current using soluble redox
mediators (Choi et al. 2003; Rabaey et al. 2007; Wang et al. 2007).
Addition of artificial exocellular redox shuttling compounds (mediators) was originally
thought to be crucial for an effective anodic electron transfer (Bennetto and Stirling 1983;
Tanisho et al. 1989). However, many recent studies have already shown that bacteria are able to
self-mediate such electron transfer either by directly using their outer-membrane redox
components (e.g. c-type cytochromes) or indirectly using self-excreted soluble redox
compounds (i.e. mediators) (Lovley 2006a; Rabaey et al. 2007; Reguera et al. 2006; Schroder
2007).
This chapter aims to establish an anodophilic biofilm from a mixed culture activated
sludge using a computer-pH controlled two-chamber microbial fuel cell (MFC). No artificial
mediator was provided during the acclimation process. The performances of a ferricyanide-
cathode and a dissolved oxygen based cathode are also compared.
Chapter 2: MFC Start-up and Basic Understandings
- 30 -
2.2 Experimental Section
2.2.1 Microbial Fuel Cell Construction
A dual-compartment microbial fuel cell made of transparent Perpex was used in the
present study (Figure 2.1, Photo 2.1). It was kindly provided by Dr. Korneel Rabaey from the
Advanced Water Management Center, The University of Queensland.
Figure 2.1 A schematic diagram of the microbial fuel cell used in the present study.
The two compartments were physically separated by a cation selective membrane
(Ultrex™ CMI7000, Membranes International Inc.) with surface area of 168 cm2. The two
compartments were of equal volume and dimension (316 mL (14 cm x 12 cm x 1.88 cm)) (Photo
2.2A). Granular graphite (El Carb 100, Graphite Sales, Inc., USA, granules 2-6 mm diameter,
porosity of 45%, see Photo 2.2B) was used as both the anode and cathode electrode, which
caused a decrease of the internal liquid volume from 316 (14 cm x 12 cm x 1.88 cm) to 120 mL
(net anodic volume). The graphite electrodes were pretreated by soaking in 1 N HCl for 24 hours,
and were washed thoroughly with running deionised water before packing into the chambers.
DAQ Board
Computer
Control/
Monitoring by
LabVIEW™
Stirrer
1M NaOH
Acetate Feed
Moisten
Air
Variable External Resistor
(PC-Controllable Relay Board)
Anolyte pH Anolyte Eh
Catholyte Eh
Ag/AgCl
Reference
Electrode
Biofilm-Anode
(Granular Graphite)
Counter Electrode
(Granular Graphite)
Cation Exchange Membrane
e- e-
Recirculation
Pump
MF
C V
olta
ge
Bio
film
-Ele
ctr
ode
Po
ten
tial
Catholyte
Graphite Rod
( Diameter =
5mm )
Out In
DAQ Board
Computer
Control/
Monitoring by
LabVIEW™
Stirrer
1M NaOH
Acetate Feed
Moisten
Air
Variable External Resistor
(PC-Controllable Relay Board)
Anolyte pH Anolyte Eh
Catholyte Eh
Ag/AgCl
Reference
Electrode
Biofilm-Anode
(Granular Graphite)
Counter Electrode
(Granular Graphite)
Cation Exchange Membrane
e- e-
Recirculation
Pump
MF
C V
olta
ge
Bio
film
-Ele
ctr
ode
Po
ten
tial
Catholyte
Graphite Rod
( Diameter =
5mm )
Out In
Chapter 2: MFC Start-up and Basic Understandings
- 31 -
Contacts between the external circuit and the granular graphite electrodes were made using
graphite rods (5mm diameter).
Photo 2.1 Photo of the microbial fuel cell set-up used in this study.
Photo 2.2 Photos of (A) the two-compartment reactor (ion-exchange membrane is not
shown) and (B) the graphite granules electrode material used in this study.
2.2.2 Bacterial Inoculum and Medium
To start up, the anode chamber was inoculated with activated sludge collected from a
wastewater treatment plant (Woodman Point, Perth, WA). 10% (v/v) of freshly collected
activated sludge with biomass concentration of 2.0 g L-1
(predetermined by using a dry-weight
method) was mixed together with synthetic wastewater. Composition of the synthetic
1 cm
B A
Chapter 2: MFC Start-up and Basic Understandings
- 32 -
wastewater was (mg L-1
): NH4Cl 125, NaHCO3 125, MgSO4·7H2O 51, CaCl2·2H2O 300,
FeSO4·7H2O 6.25, , and 1.25 mL L-1
of trace element solution, which contained (g L-1
):
ethylene-diamine tetraacetic acid (EDTA) 15, ZnSO4·7H2O 0.43, CoCl2·6H2O 0.24,
MnCl2·4H2O 0.99, CuSO4·5H2O 0.25, NaMoO4·2H2O 0.22, NiCl2·6H2O 0.19, NaSeO4·10H2O
0.21, H3BO4 0.014, and NaWO4·2H2O 0.050. Sterile concentrated yeast extract solution was
periodically added to the anolyte to support healthy microbial growth (50 mg L-1
). 100 mM of
phosphate buffer was also added to the anolyte in order to resist pH fluctuation and to improve
ionic strength of the anolyte.
2.2.3 Start-Up and Operation of a Ferricyanide-Cathode MFC
The MFC was operated in a fed-batch mode with both catholyte and anolyte
continuously re-circulating over the cathode and anode, respectively at a fixed flow rate of 8 L h-
1 (unless otherwise specified). Sodium acetate was used as the sole energy source for bacteria.
For replenishment of electron donors in the MFC, a designated amount of a concentrated sodium
acetate stock solution (1M) was injected into the anode chamber via a septum-secured injection
port just before the inlet of the anode chamber. Over the start up period, the anolyte (i.e. the
bacterial medium) was completely renewed in about every 3 days in order to remove suspended,
plank-tonic bacterial cells and any accumulated toxic end products from the system. Since the
present study is focused on the anodic process of MFC, a strong and stable cathodic reaction is
essential to maintain a strong oxidizing anode. The external resistance of the MFC was
predominately maintained at 1-5 ohm. Hence, potassium ferricyanide (50 - 100 mM),
K3Fe(CN)6 (Sigma-Aldrich, Inc., Purity ca. 99 %) was selected as the terminal electron acceptor
in the catholyte despite its non-sustainable nature compared to oxygen (air)-cathode (Freguia et
al. 2007a; Logan et al. 2006). The cathode chamber was continuously fed with 50 mM air-
saturated potassium ferricyanide, K3Fe(CN)6 solution with 100mM phosphate buffer to resist pH
changes (pH was adjusted to 7 with 5N NaOH). The ferricyanide solution was recycled over the
cathode and was running through a 2 L aeration glass bottle which contained a gas diffuser for
continuous purging of the solution with 0.45 µm-filtered airs at a flow rate of 1 L air min-1
. The
catholyte was renewed periodically to maintain a high and stable oxidation-reduction potential
for effective reduction reaction of the MFC. Loss of oxidizing power of the ferricyanide solution
is indicated by a decolorization of the solution (Photo 2.3). Designated flow rates in both
chambers were adjusted by using peristaltic pumps. Unless otherwise stated, the MFC was
operated at about 30oC and atmospheric pressure.
Chapter 2: MFC Start-up and Basic Understandings
- 33 -
Photo 2.3 Loss of oxidizing power of the ferricyanide solution indicated by a de-colorization of
the solution (left: spent solution; right: fresh solution).
Control and monitoring of the microbial fuel cell was partially automated. Potential
differences between anode and cathode (i.e. voltage), Eh in both anolyte and catholyte, and pH
of anolyte were monitored continuously by using LabVIEW™ 7.1 software interfaced with a
National Instrument™ data acquisition card and all the data were logged into an excel
spreadsheet with a personal computer (data logging frequency was depending on the signal
changes at different experiments). Since acetate oxidation always resulted in the production of
protons which caused a decrease in pH to the levels that inhibited microbial activity as reflected
by decreased current production in the MFC (data not shown), pH of the anolyte was controlled
at 7 by adding 1-2 M NaOH using a computer-controlled peristaltic pump throughout the whole
study.
2.2.4 Start-up of Anodophilic Activity in a Potentiostatic-Coupled MFC
A separate experiment was conducted to verify results obtained from the startup of
anodophilic activity with the ferricyanide-cathode, the MFC was coupled with a potentiostat
such that any possible cathodic or ohmic limitations in the system would be eliminated. The
current generation would only depend on the anodic microbial oxidation process. A three-
electrode potentiostat was used to control the anode of the same MFC at a fixed potential. The
working, counter and reference electrodes of the potentiostat were connected to the anode,
cathode and the silver-silver chloride reference electrode of the MFC, respectively. The anodic
potential was controlled at -300 mV vs. Ag/AgCl throughout the experiment. This potential was
chosen because it is slightly higher than the optimal anode potential observed in the MFC
Chapter 2: MFC Start-up and Basic Understandings
- 34 -
operated with a ferricyanide cathode in our lab (data not shown). Over the period, the anolyte
was regularly renewed (once per about 3 days). Coulombic recovery of acetate oxidation by the
anodic biofilm was tested (see below for details) by adding known amount of sodium acetate
into an acetate-starved anodic half cell.
2.2.5 Performance of a Dissolved Oxygen-based Catholyte
With the established anodic biofilm, the ferricyanide catholyte was replaced with an
aerated phosphate buffer solution in order to evaluate the performance of dissolved oxygen
(DO)-based cathode. The cathodic chamber of the MFC was repacked with fresh graphite
granules to assure the absence of any residual ferricyanide in the cathode. The catholyte (total
volume of 1L) was continuously purged with humidified air (air flow rate was 1L·h-1
) in an
external aeration container to compensate the lost of dissolved oxygen after running through the
cathode chamber. The DO concentrations (DO, mmol O2 L-1
) of the catholyte at the aeration
bucket (DOin) and outlet (DOout) of the cathodic chamber were continuously monitored with two
identical DO probes (Mettler Toledo, InPro6800; 4100 PA D/O Transmitter), respectively.
The hydraulic retention time (HRTcath.) of the catholyte (27 sec, i.e. 0.0075 h) in the
cathodic chamber was obtained from the catholyte recirculation rate (here 16 L∙h-1
). The
apparent cathodic oxygen reduction rate, ORR (mmol O2∙L-1
∙h-1
) could thus be calculated
according to eq. 2-2.
.
1- 1-
2
)()hLO (mmol ORR
Cath
inout
HRT
DODO (eq. 2-2)
2.2.5.1 Effect of Catholyte pH and Phosphate Buffer Concentration
The effect of catholyte pH on the open circuit voltage (i.e. electromotive force) was
investigated. The anode of the MFC was saturated with 10 mM acetate and the anolyte pH was
maintained at 7 by NaOH dosing. A series of phosphate buffer (50mM) solutions at pH ranging
from 0.5 to 11 were loaded into the cathodic chamber. Steady state performance of the MFC was
recorded at each pH level.
Chapter 2: MFC Start-up and Basic Understandings
- 35 -
The effect of phosphate buffer concentration of the catholyte on MFC performance was
evaluated with the DO-based cathode. A series of catholyte solutions with different phosphate
buffer concentrations (from 0 to 200 mM) was prepared (pH 7). These buffer solutions had ionic
strength ranged from 0 to 1200 mM. The ionic strength (Ic) at each buffer concentration was
calculated according to the following equation (eq. 2-3):
2
12
1i
n
i
ic ZCI
(eq. 2-3)
Where Ci is the molar concentration (mM) of ion i, Zi is the valence of that ion species, and the
sum is taken over all ions in the solution. The relationship between ionic strength of the
catholyte and the phosphate buffer concentrations is shown in Figure 2.2.
Figure 2.2 Dependency of catholyte ionic strength on phosphate buffer concentration in the
catholyte (pH 7).
y = 6.00x - 0.00
R2 = 1.00
0
200
400
600
800
1000
1200
0 50 100 150 200
Phosphate Buffer Concentration (mM)
Ionic
Str
ength
(m
M)
Chapter 2: MFC Start-up and Basic Understandings
- 36 -
2.2.6 Calculation and Analysis
2.2.6.1 Determination of Voltage, Current and Power Generation
MFC power output (P) was calculated from the voltage measured across a known
resistor (adjusted by using a variable resistor box). The current (I) was calculated according to
Ohm‘s Law, I = V/R, where V is the measured voltage and R (Ω) is the external resistance.
Power output was calculated according to: P=V x I. Since it was impossible to determine the
projected surface area of the granular graphite electrode used in this study, current and power
density could only be calculated by normalizing the values by the working volume of the anodic
chamber (i.e. 150 mL). Internal resistance (Rint) of the MFC was determined according to the
polarization slope method reported elsewhere (Logan 2008) (Refer to Appendix 2 for details)..
Coulombic efficiency was calculated by dividing the number of electrons transferred to
the anode to generate current by the number of electrons contained in the amount of acetate
being oxidized in the MFC (determined by GC-FID).Current (in milliamperes, mA) was
converted to electrons recovered by using the following conversions: 1 Coulomb (C) = 1 Amp x
1 second, 1 C = 6.24 x 1018
electrons, and 1 mol = 6.02 x 1023
electrons (96,485 C/mol).
2.2.6.2 Polarization Curve Analysis
The MFC performance was examined by conducting current-voltage analysis or the so-
called polarization curve analysis in the first month (at day 5, 14, 22 and 30). This technique is
commonly used in many other studies for generating polarization curve analysis (Logan et al.
2006). This was done by a manual step decreasing of the external resistance. Before the analysis,
the anolyte was saturated with acetate (20 mM) and a fresh ferricyanide solution was used as the
catholyte. The MFC was allowed to equilibrium at open circuit for about 2 hour to obtain the
maximal cell voltage. Thereafter, the external resistance was decreased in a step-wise manner to
1 ohm. Only steady state parameter values as defined by Logan et al. (2006) were taken to
generate the final plots. Volumetric current and power densities were obtained by normalizing
the current and power with the net anodic volume.
2.2.6.3 Acetate Analysis
A Varian Star 3400 gas chromatograph (GC) fitted with a Varian 8100 auto-sampler was
used to quantify the acetate concentration of liquid samples collected from the anodic chamber.
Chapter 2: MFC Start-up and Basic Understandings
- 37 -
The samples were subjected to centrifugation at 13,500 rpm for 5 min to separate any suspended
solids from the liquid phase. The supernatant was acidified with 1% (v/v) formic acid before
being injected (1 µL) into an Alltech ECONOCAPTM
ECTM
1000 (15 m × 0.53 mm, 1.2 µm i.d.)
column. The carrier gas (N2) was set at a flow rate of 5 mL min-1
. The oven temperature was
programmed as follows: initial temperature 80 oC; temperature ramp 40
oC min
-1 to 140
oC, hold
for 1 min; temperature ramp 50 oC min
-1 to 230
oC hold for 2 min. Injector and detector
temperatures were set at 200 and 250 oC respectively. The peak area of the Flame Ionisation
Detector (FID) output signal was computed via integration using STAR Chromatography
Software (© 1987-1995).
Chapter 2: MFC Start-up and Basic Understandings
- 38 -
2.3 Results and Discussion
2.3.1 Quick Start-up of MFC using Ferricyanide-Cathode and Continuous pH-Static
Control
The described two-chamber MFC was inoculated with activated sludge, and operated for
30 days using ferricyanide as the catholyte. Over this acclimation period, the external resistance
of the MFC was predominately maintained at 1-5 ohms.
Polarization curve analysis was conducted regularly over time (Figure 2.3). It allows the
determination of both the maximal current and power densities as well as the electrochemical
properties of the anode and cathode. In general, the volumetric current and power densities
(normalized to net anodic chamber volume) of the MFC increased over time (Figure 2.3), and
the internal resistance of the MFC had decreased over time, reaching the lowest internal
resistance of only 0.54Ω at day 30 (Table 2.1). With a similar MFC set-up, Dr. Korneel Rabaey
(Advanced Water Management Center, The University of Queensland) had also recorded a
similar internal resistance of 0.61Ω by using electrochemical impedance spectroscopy method
(personal communication on 14th
, Sep 2007).
The overtime decrease in internal resistance is indicative for a well performing
anodophilic microbial community and demonstrates the effective design and operation of the
reactor. The bacterial culture in the anode chamber has become more capable in transferring
their electrons to the anode during the acclimation.
Table 2.1 A summary of the MFC performance over a 30-day development of the anodophilic
biofilm in the ferricyanide cathode-MFC.
Time Maximal Power
Density Maximal Acetate Oxidation Rate
External Resistance at Maximal Power
Density Internal Resistance*
Day W·m-3
mM·h-1
Ω Ω
5 415 1.2 6 2.26
14 868 3.0 2 1.20
21 1321 5.2 1 0.97
30 1735 6.1 1 0.54
Note: the maximal acetate oxidation rates are calculated assuming a 100% Coulombic conversion
of acetate into current (8 e- mole per acetate); *internal resistances were calculated by using
polarization slope method (Logan 2008, page 53-54) (see Appendix 2).
Chapter 2: MFC Start-up and Basic Understandings
- 39 -
Figure 2.3 The power density (A) and polarization curves (B, cell voltage; C, electrode potential)
of the MFC at day 5 ( / ), day 14 ( / ), day 22 ( / ) and day 30 ( / ) after
inoculated with activated sludge (10%, v/v).
0
200
400
600
800
Cell
Voltage
(mV
)
0
500
1000
1500
2000
Pow
er
Density
(W m
-3,
Anodic
Void
Vol.)
.
-600
-400
-200
0
200
400
0 1000 2000 3000 4000
Current Density (A m-3
, Anodic Void Vol.)
Ele
ctr
ode P
ote
ntial
(mV
vs.
Ag/A
gC
l)
A
B
C
Chapter 2: MFC Start-up and Basic Understandings
- 40 -
The power density obtained at day 5 (415 W·m-3
) has already approached the maximal
levels observed in other MFCs that have been operated for a longer acclimation period
(Aelterman et al. 2008; Kim et al. 2007). For instance, with a similar reactor geometry, ion-
exchange membrane and electrode materials, Aelterman et al. (2008) observed a maximal MFC
power density of 517 W·m-3
(also normalized to the net anodic volume) but after a 25 days
acclimation period. Instead of activated sludge, an effluent taken from an active acetate-fed
MFC was used as their start up inoculum. Overall, the result obtained here suggests that quick
establishment of anodophilic activity from an activated sludge could be achieved by using a
ferricyanide-cathode and continuous neutral pH static control.
2.3.1.1 The Cathodic Oxidation Power could influence the Anodic Microbial Activity only
at a Low External Resistance
The use of a ferricyanide-cathode had offered a strong oxidizing surface (anode) to the
electrochemically active bacteria for the efficient oxidation of acetate during the acclimation.
However, the rate of acetate degradation related strongly to the oxidation power of the catholyte
(i.e. catholyte Eh) only when the external resistance was low (here at 1 Ω) (Figure 2.4).
Figure 2.4 Effect of ferricyanide-catholyte redox potential (Eh) on the anodic acetate oxidation
rate at different external resistances. Acetate oxidation are calculated assuming a 100%
Coulombic conversion of acetate into current (8 e- mole per acetate). In all cases, the
anode was always saturated with acetate (10 mM) and catholyte pH remained at 7.
0
1
2
3
4
5
0 50 100 150 200
Catholyte Eh (mV vs. Ag/AgCl)
Ace
tate
Oxid
atio
n R
ate
(mM
·h-1
)
1Ω
5Ω
10Ω
100Ω
1MΩ
Weak Oxidant
Strong Oxidant
Chapter 2: MFC Start-up and Basic Understandings
- 41 -
Such correlation seems to obey with the well known Michaelis-Menten type kinetics
behavior: where increasing oxidation power of the catholyte (catholyte Eh) increases the
substrate oxidation rate up to a point where further increase in catholyte Eh could not further
increase the substrate oxidation rate (Figure 2.4, 1 and 5 Ω). While at external resistances of 10
Ω or higher, increase in the oxidation power of the ferricyanide catholyte had no significant
effect on the anodic acetate oxidation rate.
Similar observation has been reported by You and coworkers (2006). With low
resistance between anode and cathode, increase in concentrations of the terminal electron
acceptors (permanganate was used in their study) in the cathodic chamber of the MFC could
increase current density whereas no significant effect was observed when higher resistance was
used.
Overall, the use of a very low external resistance (less than 5 Ω in the described MFC)
was beneficial or perhaps crucial to enable effective anodic oxidation by the electrochemically
active bacteria during the MFC operation.
2.3.2 Potentiostatically-Controlled MFC also revealed Quick Anodophilic Activity Onset
from the same Activated Sludge Inoculum
If the observed fast establishment of anodophilic activity was attributed to the provision
of a strong oxidizing surface (anode) to the bacteria (via ferricyanide with low external resistor),
other methods that can maintain a similar oxidizing capacity of the anode should also allow a
quick start up of anodophilic activity from the activated sludge under similar anodic condition.
One of these methods is by coupling the MFC with a potentiostat. Using potentiostats for
the evaluation of anodic processes in BES has been proposed as an appropriate method to rule
out system limitations such as ohmic resistance and cathodic limitations (Fricke et al. 2008;
Marken et al. 2002). Hence, a potentiostatic experiment was conducted to establish anodophilic
activity from the same activated sludge inoculum under a similar anolyte condition (pH 7, 30oC).
The anodic potential was controlled at a constant level (here -300 mV vs. Ag/AgCl). The
ferricyanide at the cathode was replaced with a 50 mM phosphate buffer (pH7) (Figure 2.5).
Chapter 2: MFC Start-up and Basic Understandings
- 42 -
Figure 2.5 Evolution of (A) current and (B) anolyte Eh after inoculation of an activated sludge
(10% v/v) in a pH-controlled potentiostatic bioelectrochemical cell (the same as used in
Figure 2.2). Initial acetate concentration in the medium was about 15 mM. pH was
controlled at about 7 by feedback controlled addition of a 1M NaOH solution (C). Values
reported in A and B are the means taken in 30 min intervals. Dotted arrows indicate
additions of 50mM sodium acetate.
2.3.2.1 Evolution of Anodic Current after a short Lag-phase of only One Day
After the bacterial inoculation, the anolyte Eh decreased almost linearly from about -50
to a steady level of about -430 mV in 20 hours (Figure 2.5B). Such a decline in Eh indicates that
the activated sludge could reduce their surrounding environment (anolyte) to a highly reduced
-500
50100150200250300350400450500550
Curr
ent (m
A)
-450
-350
-250
-150
-50
An
oly
te E
h (
mV
vs. A
g/A
gC
l) .
4
7
10
0 1 2 3 4 5 6 7 8 9Time (d)
pH
A
B
C
Medium renewal
Medium renewal
CE = 95.7%
(1)
(2)
(3)
(4) (5)
(6)
Chapter 2: MFC Start-up and Basic Understandings
- 43 -
condition by metabolizing the acetate (electron donor). Such anaerobic condition would be
essential for the effective anodic electron transfer by the bacteria.
After a lag phase of only one day anodic current began to evolve exponentially,
signifying the onset of exponential growth of anodophilic active microorganism at the anode
(Figure 2.5A) (Hu 2008; Parot et al. 2007). Upon acetate depletion (data not shown), the anodic
current returned to the background, low level. This result demonstrates the ease of establishing
an anodophilic biofilm from an initially electrochemically inactive activated sludge inoculum.
2.3.2.2 Anodophilic Bacteria: Biofilm instead of Suspended Cells
At about day 3, the entire anolyte was discarded and the anodic chamber was flushed
with one liter (about 6 times the net anodic chamber volume) of fresh medium to remove any
plank-tonic cells or suspended cell debris before resuming the operation with an acetate-
saturated (50 mM) fresh medium (Figure 2.5A). Such flushing treatment did not retard the
anodophilic activity of the bacteria, but indeed a sharp increase in anodic current was observed
almost immediately after the medium renewal. Another flushing and medium renewal at about
day 7 resulted in a similar effect on anodic current (Figure 2.5A). Hence, the observed anodic
current was not due to the suspended bacteria but to the metabolizing biofilm attached to the
anode. This observation has also been reported by others using activated sludge as the source of
anodophilic bacteria (Lee et al. 2006; Yoon et al. 2007).
2.3.2.3 The Anodophilic Biofilm Established in the Potentiostatically Controlled MFC
also gave similar Power Output as the Ferricyanide-Cathode MFC
To compare the anodophilic biofilms established from the ferricyanide-cathode MFC and
the potentiostatically controlled MFC on a power production basis, the current obtained from the
latter was further processed to obtain an extrapolated hypothetical power output (Figure 2.6).
Assuming that the biofilm-anode was coupled to a ferricyanide-cathode, the theoretical MFC
cell voltage can be deduced from the difference between the poised anodic potential (here -300
mV vs. Ag/AgCl) and the cathodic potential at the corresponding current level. With this
estimated voltage, the extrapolated power output was calculated according to Power = Voltage ×
Current.
Chapter 2: MFC Start-up and Basic Understandings
- 44 -
Figure 2.6 Extrapolated MFC power output using the anodophilic biofilm established from a
potentiostatic cell. Three power outputs values are depicted here. They are obtained from
peak 1, 3 and 5 in Figure 2.5. Data of the ferricyanide cathode are taken from Figure
2.3C (day 30). Cath. E = cathodic potential; An. E = anodic potential.
Similar to the ferricyanide-cathode MFC, the extrapolated power output of the
potentiostatically established anodophilic biofilm also increased over time (Figure 2.6). The
power outputs of the potentiostatically controlled MFC are in general higher than that of the
ferricyanide-cathode MFC. For instance, at day 5 the power densities of the ferricyanide-cathode
MFC and the potentiostat-MFC were 415 and 864 W·m-3
, respectively. This may be due to the
slightly more favorable operating anodic potential (-300 mV compared to <-400 mV))
maintained by the potentiostat. Nevertheless, it can be concluded that the anodophilic biofilms
established by using these two different approaches were similar.
Overall, this estimation suggests that even a young (ca. one week) anodophilic biofilm
could easily enable a MFC power output beyond a benchmark target of 400 W m-3
(as proposed
by Clauwaert et al. (2008a)) if both the cathodic reaction and electrolyte pH were not limiting
during the acclimation phase.
0
100
200
300
400
500
-400 -200 0 200
Electrode Potential (mV vs. Ag/AgCl)
Curr
ent
(mA
)Bioanode Potential(Poised at -300 mV)
Cathodic Potential(Ferricyanide-cathode)
Day 2.3 = 457 W·m-3
Day 5 = 864 W·m-3
Day 7.4 = 904 W·m-3
Hypothetical Power Output:
Power = (Cath.E – An.E)×Current
Chapter 2: MFC Start-up and Basic Understandings
- 45 -
Figure 2.7 Scanning electron micrographs of the granular graphite anode before (D, E and F)
and after 200 days (A, B and C) in an acetate-fed microbial fuel cell anodic chamber.
2.3.2.4 SEM reveals only Thin and Low Biomass Density at the Highly Active Biofilm-
Anode
After operating the MFC for about 200 days, scanning electron microscopy photos were
taken to examine the morphology of the biofilm attached on the granular graphite anode (Figure
2.7). Surprisingly, only a thin, open structured and low biomass density biofilm was observed at
the electrode surface. Examination of replicates samples revealed similar morphology. This
implies that the established anodophilic biofilm has a very high specific activity (current
generation per amount of biomass).
The observation here seems to corroborate with the results reported by Altermam et al.
(2008). In their active MFC anodic chambers, these authors had determined a maximal biomass
density of 0.58 g VSS∙L graphite -1
, which they claimed is approximately 30 times lower
compared to biomass concentrations in anaerobic digesters (10-20 g VSS∙L-1
) and yet the
specific biomass activity of their biofilm (3.4 g COD g∙VSS-1
∙d-1
) was higher than that of
anaerobic systems (about 2 g COD g∙VSS-1
∙d-1
) (Aelterman et al. 2008; van Lier et al. 2008)
(Table 2.2).
If we assumed the anodic chamber of the described MFC in this study also has the same
biomass density of 0.58 VSS∙L graphite -1
, the specific biomass activity obtained at the maximal
10 µm 5 µm
2 µm10 µm 5 µm
2 µm10 µm10 µm10 µm 5 µm5 µm5 µm
2 µm2 µm2 µm10 µm10 µm10 µm 5 µm5 µm5 µm
2 µm2 µm2 µm
A B C
E D F
Chapter 2: MFC Start-up and Basic Understandings
- 46 -
power output at day 30 would be 17.4 g COD g∙VSS-1
∙d-1
, which is about 6 times higher than the
typical activity of other MFC reported in the literature and is about 35 and 9 times higher than
that of activated sludge and anaerobic digestion treatment, respectively (Table 2.2) (Pham et al.
2009). This distinctive feature may give MFC a competitive advantage over other wastewater
treatment processes especially activated sludge processes as the post-treatment cost of the
excessive sewage sludge is tremendous.
Table 2.2 Comparison of parameters characterizing MFC anode, aerobic activated sludge and
anaerobic digestion for wastewater treatment.
Wastewater Treatment Technology
Yield Coefficient (Y)
Biomass Concentration (per reactor volume)
COD Removal Rate
Specific Biomass
Activity
g∙VSS-1∙g COD
-1 g∙VSS
-1∙L
-1 kg COD m
-3 d
-1 g COD g∙VSS
-1∙d
-1
a Aerobic Activated
Sludge 0.4 2 - 4 0.5 - 1 0.1 0.5
a Anaerobic Digestion 0.05 – 0.1 10 - 20 10 - 20 1 - 2
a MFC Anode
Literature 0.02 – 0.23 0.25 - 0.58 0.3 – 1.5 1.51 - 6.93
MFC Anode This Chapter
0.1 a 0.4
a 3.0 - 9.0
b 6.5 - 17.4
c
a Values after Pham et al.(2009);
b Values are estimated from current (100% Coulombic efficiency was
assumed); c Values are estimated based on a current output of the described MFC between 150 and
450 mA.
2.3.3 Performance of a Dissolved Oxygen-based Cathode
In this study, the use of ferricyanide enabled the effective start up of anodophilic activity
and the high electricity output of the MFC. However, ferricyanide is not sustainable and hence it
has very limited merit for wastewater treatment application. Oxygen, on the other hand, is
considered as the most sustainable electron acceptor in MFC cathode because of its virtually
unlimited availability from the atmosphere, and in principle it has the highest redox potential
(Freguia et al. 2007b; Oh et al. 2004). However, it is well accepted that the imperfect catalysis of
the oxygen reduction reaction (eq. 2-4) at plain, non-catalyzed electrode surface limits the
cathodic reaction and hence the MFC performance.
Chapter 2: MFC Start-up and Basic Understandings
- 47 -
O2 + 4 H+ + 4 e
- 2 H2O (eq. 2-4)
With the same established anodic biofilm, the ferricyanide catholyte was replaced with
an aerated phosphate buffer solution to evaluate the performance of a dissolved oxygen (DO)-
based cathode. In particular, the effects of catholyte pH and phosphate buffer were evaluated.
2.3.3.1 Acidified Catholyte increases Open Circuit Voltage of the MFC
Acidifying the catholyte (50 mM aerated phosphate buffer) could increase the open
circuit voltage (OCV) of the MFC (Figure 2.8A). Such effect was due to an increase in the
cathodic potential (Figure 2.8B). From a thermodynamic perspective, the electrode potential in a
MFC is linearly dependent on the local pH of the electrolyte (refer to Figure 1.8 in Chapter 1).
Decrease in catholyte pH by one unit should increase the cathodic potential by 59 mV. However,
such dependence did not hold true here, especially at a higher pH range (>pH7). For example,
when the catholyte pH dropped from about 11 to 9.3, the catholyte potential increased from -67
to -64 mV (2 mV/pH). When the catholyte pH changed from 3 to 0.5, the catholyte potential
increased from +170 to +360 mV (76 mV/pH) (Figure 2.9).
Figure 2.8 (A) Plots of cell voltage; cathodic potential and anodic potential as a function of
catholyte pH under open circuit operation. (B) Correlation between the open circuit
potential and the cathodic potential in A. All catholyte contained 50 mM phosphate
buffer. All values were recorded at equilibrium state. Anolyte was saturated with 5 mM
acetate, pH 7. The ferricyanide cathode data in A was obtained from Figure 2.3B.
Catholyte pH
Ce
ll V
olta
ge
(m
V) o
r E
lectr
od
e
Po
ten
tia
l (m
V v
s. A
g/A
gC
l)
-600
-400
-200
0
200
400
600
800
0 2 4 6 8 10 12
Cell Voltage
Cathodic Potential
Anodic Potential
OCV of Fe(CN)63-Cathode
y = 1.009x + 520.940
R2= 0.999
0
200
400
600
800
-100 0 100 200 300 400
Cathodic Potential (mV vs. Ag/AgCl)
Op
en
Circu
it P
ote
ntia
l (m
V)
(A) (B)
Chapter 2: MFC Start-up and Basic Understandings
- 48 -
Figure 2.9 Change in cathodic potential per unit of pH as a function of the catholyte pH.
(plotted from the data in Figure 2.8A).
Similar observation was noted twice by Zhao and coworkers from their experiments
using graphite electrodes modified with various chemical catalysts. They attributed the
phenomena to a change in the nature of the rate-determining step in the overall cathodic reaction,
most likely from a 4 e-/ 4H
+ (or 2 e
-/ 2H
+) to a 4 e
-/ 2H+ (or 2 e
-/ 1H
+) reaction (relative to one
oxygen molecule) (Zhao et al. 2005, 2006a).
Whether or not the observed discrepancy was due to the same reason is beyond the scope
of the present study. Yet, the significance of the result here is that when compared with a
ferricyanide cathode, which always gave an OCV of about 750 mV the oxygen-based cathode
could give similar OCV only at catholyte pH of less than 2 (Figure 2.8A). Such an acidic
condition is unlikely to be realistic under normal MFC operation, especially in systems designed
for wastewater treatments. On the other hand, the cathodic reduction of ferricyanide does not
involve protons (eq. 2-5), and hence varying the catholyte pH does not affect the cathodic
potential (Zhao et al. 2006b). Based on this (but ignore the sustainability concern), ferricyanide
can allow a more stable cathodic reduction reaction for the MFC compared to a DO based
cathode.
Fe(CN)63-
+ e- → Fe(CN)6
4- (eq. 2-5)
2.3.3.2 Phosphate Buffer improves the Dissolved Oxygen-Cathode Performance
Increasing the phosphate buffer concentration in the catholyte could increase the
0
20
40
60
80
0 2 8 10 12Catholyte pH
Cath
odic
Pote
ntial C
hange
mV
/pH
Theoretical Change
59 mV/pH
Chapter 2: MFC Start-up and Basic Understandings
- 49 -
performance of the DO-cathode (Figure 2.10). When DO was not limiting (>5 mg L-1
), a
Monod-type relationship between the MFC current and the buffer concentration was observed:
maximal current ≈ 40 mA; half saturation buffer concentration ≈ 8 mM, and saturation buffer
concentration ≈ 100 mM (Figure 2.10A). This result indicates that the cathodic oxygen reduction
depends on the phosphate buffer concentration in the catholyte (Figure 2.10C).
Figure 2.10 Effect of phosphate buffer concentration in catholyte on (A) voltage; current; power
density; and (B) electrode potentials of the acetate-fed microbial fuel cell; (C) cathodic
oxygen reduction rate as a function of phosphate buffer concentration in the catholyte.
External resistance 1.5 Ω. Anolyte was saturated with about 5 mM acetate. pH of both
anolyte and catholyte was 7, temperature 30oC. Catholytes were always saturated with
>5 mg·L-1
dissolved oxygen.
The positive effect of phosphate buffer was possibly due to its capacity to improve the
proton availability for the cathodic reaction, which consumes protons (eq. 2-4). At pH 7, proton
concentration is very low, i.e. 0.001 mM. With a poorly pH-buffered catholyte, proton
consumption at the cathode would likely be faster than the proton diffusion from the bulk to
compensate for the proton loss at the cathode. Consequently a localized pH increase is
0
10
20
30
40
50
60
70
Cell Voltage
Current
Power Density
-520
-500
-480
-460
-440
0 200 1000 1200
Cathode
Anode
(B)
(A)
0 50 100 200150
Ionic Strength (mM)
Vo
lta
ge
(m
V);
Cu
rre
nt (
mA
);
Po
we
r D
en
sity (W
·m-3
)
Ele
ctr
od
e P
ote
ntia
l(m
V v
s. A
g/A
gC
l)
Phosphate Buffer (mM)
0 200 1000 1200Ionic Strength (mM)0
1
2
3
4
5
Oxyg
en R
eduction R
ate
(mm
ol∙O
2∙L
-1∙h
-1)
0 50 100 200150
Phosphate Buffer (mM)
(C)
Chapter 2: MFC Start-up and Basic Understandings
- 50 -
established at the cathode, shifting the cathodic potential to a more negative value and hence the
cell voltage and current are diminished (Zhao et al. 2006b).
The situation would be very different if the bulk catholyte contained a sufficient amount
of phosphate buffer. For instance, at a phosphate buffer concentration of 100 mM, the diffusion
rate of the phosphate buffer species in the catholyte could be about 105 times higher than that of
protons, even though the diffusion coefficient of free protons is about 10 times higher than that
of buffering species (monobasic and dibasic phosphate: 1.0 × 10-5
and 0.86 × 10-5
cm2·s
-1 at
30oC (Fan et al. 2007; Mason and Culvern 1949)). Hence, the buffer-mediated transfer of
protons from the bulk catholyte to the cathodic reaction site could be efficient enough to meet
the proton demand in the cathode oxygen reduction, resulted in an improved cathode and MFC
performance.
Nevertheless, there are at least two drawbacks that may disfavor the use of phosphate
buffer in BEC for wastewater treatment: 1) huge amount of buffer is required especially for
large scale wastewater treatment, production and energy cost of the buffer would be huge; 2)
discharge of the buffer-containing effluent would potentially cause subsequent pollution at the
receiving water bodies, e.g. eutrophication. Hence, a more sustainable alternative of
compensating the proton demand of the oxygen-based cathodic reaction has to be developed.
2.3.3.3 Ferricyanide-Cathode Outperforms Dissolved Oxygen-Cathode
In spite of the beneficial effect of phosphate buffer on the cathodic oxygen reaction, the
maximal current of the DO-cathode MFC was only about 40 mA, which was about 8 times
lower compared to the current obtained from the ferricyanide-cathode MFC (Figure 2.11).
The inferior performance of the DO-cathode MFC was largely due to the huge
overpotential of the cathode (Freguia et al. 2007b; Rismani-Yazdi et al. 2008). The cathodic
potential of the DO-cathode was about 12 times lower (more negative) compared to ferricyanide
cathode (Figure 2.11), suggesting that the cathodic oxygen reduction is associated with a large
overpotential compared to ferricyanide.
Similar to others, the abiotic cathodic oxygen reduction was the limiting step to the
current generation in the described MFC (Clauwaert et al. 2007b; Freguia et al. 2007b; Oh et al.
2004; Rismani-Yazdi et al. 2008). Further studies are required to improve the cathodic oxygen
reduction in a sustainable manner.
Chapter 2: MFC Start-up and Basic Understandings
- 51 -
Figure 2.11 Comparison between ferricyanide- and dissolved oxygen- based cathode in an
acetate-fed microbial fuel cell. External resistances of both settings were 1.5 Ω. Anolyte
was saturated with 5 mM acetate, pH7, 30oC. Both catholytes were amended with 100
mM phosphate buffer (pH7). Data of the ferricyanide cathode are taken from Figure 2.3C
(day 30). CP/ AP = cathodic/ anodic potential (mV vs. Ag/AgCl).
2.4 Conclusion and Implication
This chapter demonstrates that quick establishment of anodophilic activity from an
activated sludge could be achieved by using a ferricyanide-cathode and continuous neutral pH
static control of the MFC anode. Both the volumetric current and power densities of the MFC
increased over time. The overtime decrease in internal resistance is indicative for a well
performing anodophilic microbial community and demonstrates the effective design and
operation of the reactor. The bacterial culture in the anode chamber has become more capable in
transferring their electrons to the anode during the acclimation period. The power density
obtained at day 5 (415 W·m-3
) had already approached the maximal levels observed in other
similar-scale MFCs that have been operated for a longer acclimation period.
The developed anodophilic biofilm appeared to have a very high specific activity in
converting the soluble COD (here acetate) into electricity. This distinctive feature may give
MFC a competitive advantage over other wastewater treatment processes especially activated
sludge processes as the post-treatment cost of the excessive sewage sludge is tremendous.
-500
0
500
1000
1500
Current
(mA)
CP(mV) AP (mV)
Cell
Voltage(mV)
Power
Density(W·m-3)
Ferricyanide
Dissolved Oxygen
Chapter 2: MFC Start-up and Basic Understandings
- 52 -
As demonstrated in this study, ferricyanide (Fe (CN)63-
) was a very effective terminal
electron acceptor for the cathodic process. It offered a stable reduction reaction even at the non-
catalyzed granular graphite cathode surface. However, due to the slow reoxidation of the
reduced state (ferrocyanide, Fe (CN)64-
), the regular catholyte renewal was essential to sustain
stable MFC performance. This requirement leads to the generation of a tremendous amount of
spent non-functioning ferrocyanide solution (chemical waste), leading to a cost and
environmental impact of the process.
Oxygen, on the other hand, is considered as the most sustainable electron acceptor in
MFC cathode because of its virtually unlimited availability from the atmosphere. It was
demonstrated that the addition of phosphate buffer could enhance the cathodic oxygen reduction,
most likely due to a facilitated buffer-mediated proton transfer from the bulk catholyte to the
cathode. However, the performance of the dissolved oxygen based cathode was insignificant
compared to that of a ferricyanide-cathode. Hence, future study should aim at improving the
oxygen-based cathodic reaction. Lastly, the continuous pH control by NaOH addition and the
regular replenishment of catholyte used in this study may facilitate the ion transfer in the system
in a non sustainable manner, allowing Na+ instead of H
+ transfer as the dominant mechanism of
ionic charge transfer. From a practical standpoint, a more sustainable pH control method has to
be developed.
- 53 -
3 Affinity of Microbial Fuel Cell Biofilm for the
Anodic Potential
(This chapter has been published in Environ. Sci. Technol. (2008) 42: 3828-3834)
Chapter Summary
In Chapter 2, the relationship between the anodic acetate oxidation rate and the catholyte
oxidation power (catholyte Eh) seemed to exhibit a well known Michaelis-Menten type kinetics
behavior: where at low external resistance (<5Ω) increasing catholyte Eh increases the substrate
oxidation rate until reaching a point beyond which further increase in catholyte Eh could not
further increase the substrate oxidation rate.
In analogy to the well established dependency of microbial reactions on the redox
potential of the terminal electron acceptor, the dependency of the microbial activity in a highly
active microbial fuel cell on the potential of the electron accepting electrode (anode) in a
microbial fuel cell (MFC) is investigated in this chapter. An acetate-fed, pH controlled MFC
was operated for over 200 days to establish a highly active MFC anodic biofilm using
ferricyanide as the catholyte and granular graphite as electrode material. From the coulombic
efficiency of 83% of the MFC the microbial activity could be recorded by on-line monitoring of
the current.
The results suggest that:
(1) in analogy to the Michaelis-Menten kinetics a half saturation anodic potential (here
termed kAP value) could be established at which the microbial metabolic rate reached
half its maximum rate. This kAP value was about -455 mV (vs. Ag/AgCl) for our
acetate driven MFC and independent of the oxidation capacity of the cathodic half cell;
(2) a critical AP (here termed APcrit.) of about -420 mV (vs. Ag/AgCl) was established that
characterizes the bacterial saturation by the electron accepting system. This critical
potential appeared to characterize the maximum power output of the MFC.
This information would be useful for modeling and optimization of microbial fuel cells and
the relative comparison of different microbial consortia at the anode.
Chapter 3: Affinity of MFC Biofilm for the Anodic Potential
- 54 -
3.1 Introduction
Microbial fuel cell technology (MFC) is an emerging environmental technology for
energy recovery from organic waste streams. Over the past few years, our knowledge on MFC
has been gaining momentum. Yet our understanding of the electron transfer from the bacteria to
the anode is still not complete (Kim et al. 2007; Lovley 2006b; Rabaey et al. 2007; Schroder
2007). In MFC an electron flow is generated between the anodic half cell in which the bacteria
form a biofilm on the anode surface and the cathodic half cell which uses a chemical electron
sink such as oxygen or ferricyanide (Logan et al. 2006). In the anodic half cell bacteria oxidize
organic electron donors such as acetate via the tricarboxylic acid (TCA) cycle and transfer the
gained electrons to internal electron carriers such as NAD+/NADH. A suitable external electron
acceptor is needed for the bacteria to recycle (re-oxidise) their internal electron carriers. Such an
acceptor may either be present extracellularly in the environment or generated intracellularly by
the bacteria (De Graef et al. 1999; Stams et al. 2006). In MFC the bacteria manage to transfer
the internally generated electron flow to a suitable solid conductive surface (anode). As the
anode receives the electrons its potential decreases to a lower level than that of the cathode in
the adjacent cathodic half cell. If a suitable resistor connects the anode and cathode an electron
flow can be measured that is generated by the bacteria. In nature, such a microbially initiated
electron flow normally ends by the transfer of electrons to the terminal electron acceptor such as
oxygen, nitrate, sulfate or even poorly soluble minerals (Clauwaert et al. 2007a; Newman and
Kolter 2000; Rabaey et al. 2005b; Zhao et al. 2006a). It has been shown that electron acceptor
couples with a more positive redox potential enable bacteria to oxidize the electron donor more
effectively as it provides a greater amount of Gibbs Free Energy Change (Cord-Ruwisch et al.
1988).
It is conceivable that also the bacteria in MFC profit from a more positive potential of
the external electron acceptor (here the anodic potential), as this increases the energetics (more
negative Gibbs free energy change) of the bacterial reaction. From a biological point of view one
would expect that very negative anodic potential (the anodic potential approaching that of the
electron donor (here acetate)) limits the microbial capacity to pass electrons to the anode.
Further one would expect that at very positive anodic potentials the microbial activity is limited
by factors other than the availability of a suitable electron accepting system and that ―electron
acceptor saturation‖ will be observed, similar to the substrate saturation known for enzymatic
and microbial processes.
Chapter 3: Affinity of MFC Biofilm for the Anodic Potential
- 55 -
MFC offers the possibility of varying the potential of the electron acceptor (here the
anode) without changing the reacting species (e.g. nitrate, oxygen, sulfate), and hence a MFC is
a suitable bio-electrochemical tool to study this fundamental aspect of microbial physiology
(Rabaey et al. 2007). In the literature, the dependency of the microbial metabolic activity on the
anodic potential of a MFC has only been tested with only limited detail and no quantitative
study has been published (Schroder 2007). In a recent study on a benthic bio-electrochemical
system the anodic potential was controlled artificially by a potentiostat and the authors could
show that there was a tendency of increasing microbial activity with increasing anodic potentials
(Finkelstein et al. 2006).
It is the objective of this study to establish a generic understanding of the behavior of
anodophilic bacteria in a MFC when the anodic potential becomes limiting. Similar to the
widely used saturation behavior for soluble species, such knowledge is expected to assist with
evaluating the efficiency of microbial populations in MFC. In order to test the undisturbed
activity, in-situ, we used the anodic potentials as they were generated and changed by the
bacterial biofilm in the MFC. Different approaches were adopted to obtain the relationship
between microbial activity and anodic potential in the MFC. These included: (1) manual step
changes of external resistance as commonly used in many other studies for polarization curve
analysis; (2) constant sweeping of external resistance and (3) open circuit to allow building up of
anodic potential under substrate saturated or limited conditions. The interrelationship among
microbial activity, anodic potential and MFC power output is explored in detail. Further it is
attempted to explain the maximum power output of the fuel cell by the microbial capacity to
discharge electrons to the anode at different anodic potentials.
Chapter 3: Affinity of MFC Biofilm for the Anodic Potential
- 56 -
3.2 Experimental Section
3.2.1 Microbial Fuel Cell Start-up and On-line Process Monitoring
A dual-compartment microbial fuel cell made of transparent Perspex was used in the
present study. The two compartments were physically separated by a cation selective membrane
(CMI-7000, Membrane International Inc.) with a surface area of 168 cm2. The two
compartments were of equal volume and dimension (316 mL (14 cm x 12 cm x 1.88 cm)).
Conductive granular graphite (El Carb 100, Graphite Sales, Inc., USA, granules 2-6 mm
diameter) was used as both the anode and cathode electrodes, which decreased the internal liquid
volume from 316 to 120 mL. Contacts between the external circuit and the granular graphite
electrodes were made using graphite rods (5 mm diameter).
To start up the MFC experiment, the anode chamber was inoculated with activated
sludge collected from a local wastewater treatment plant (Woodman Point, Perth, WA,
Australia). 10% (v/v) of freshly collected activated sludge with a biomass concentration of about
2.0 g L-1
was mixed with synthetic wastewater. The composition of the synthetic wastewater
was (mg L-1
): NH4Cl 125, NaHCO3 125, MgSO4·7H2O 51, CaCl2·2H2O 300, FeSO4·7H2O 6.25,
and 1.25 mL L-1
of trace element solution, which contained (g L-1
): ethylene-diamine tetra-acetic
acid (EDTA) 15, ZnSO4·7H2O 0.43, CoCl2·6H2O 0.24, MnCl2·4H2O 0.99, CuSO4·5H2O 0.25,
NaMoO4·2H2O 0.22, NiCl2·6H2O 0.19, NaSeO4·10H2O 0.21, H3BO4 0.014, and NaWO4·2H2O
0.050. Sterile concentrated yeast extract solution was periodically added (ca. every 3 to 5 days)
to the anolyte (50 mg L-1
final concentration) as growth nutrients.
The MFC was operated in a fed-batch mode with both catholyte and anolyte
continuously re-circulating over the cathode and anode, respectively. Unless otherwise stated,
the recirculation rates were 8.0 L h-1
to minimize mass transfer limitation. Sodium acetate was
used as the sole electron donor for the bacterial culture. For replenishment of electron donors in
the MFC, a designated amount of a concentrated sodium acetate stock solution (1M) was
injected into the anode chamber via a septum-secured injection port just before the inlet of the
anode chamber.
Since the present study is focused on the anodic process of MFC, a strong and stable
cathodic reaction is essential for effective control of the anodic potential. Hence, potassium
ferricyanide (50 - 100 mM), K3Fe(CN)6 (Sigma-Aldrich, Inc., Purity ca. 99 %) was selected as
the terminal electron acceptor in the catholyte despite its non-sustainable nature compared to
Chapter 3: Affinity of MFC Biofilm for the Anodic Potential
- 57 -
oxygen (air)-cathode (Freguia et al. 2007a; Logan et al. 2006). Air-saturated ferricyanide
solution complemented with 100 mM phosphate buffer (pH 7) was continuously re-circulated
through the cathode chamber at a fixed flow rate to reduce cathodic over-potential. Outside of
the MFC the catholyte was circulated through a 2 L aerated glass bottle which contained a gas
diffuser for continuous purging of the solution with water-saturated air at a flow rate of 1 L min-
1. Unless a weaker cathodic process was desired in some experiments, the catholyte was renewed
periodically to maintain a reasonably stable cathodic potential.
Designated flow rates in both chambers were adjusted by using peristaltic pumps. Unless
otherwise stated all experiments were performed at a constant temperature (30oC) and ambient
atmospheric pressure. Control and monitoring of the microbial fuel cell was partially automated.
Potential differences between anode and cathode, Eh in both anolyte and catholyte, and pH of
anolyte were monitored continuously by using LabVIEW™ 7.1 software interfaced with a
National Instrument™ data acquisition card (DAQ). The anodic potential was measured against
a silver/ silver chloride reference electrode (saturated potassium chloride) which was placed
inside the anodic compartment. Accuracies of all the voltage signals detected by the DAQ were
regularly checked by using a high precision digital multimeter (resolution 10 µV). All data were
logged into an Excel spreadsheet with a personal computer (data logging frequency depended on
the signal changes at different experiments).
As virtually no net proton migration across the cation exchange membrane is observed
under normal MFC operating condition (Rozendal et al. 2006a), protons (H+) generated from
microbial oxidation of acetate accumulated, resulting in a pH drop of the anolyte. This pH drop
was found to inhibit microbial activity as reflected by diminished electron flow of the MFC
(data not shown) (Gil et al. 2003; Zhao et al. 2006a). To overcome the problem, the pH of the
anolyte was strictly controlled at 7 ± 0.2 by the automated addition of NaOH (1M) using a
computer-controlled peristaltic pump throughout the whole study. In experiments where acetate-
saturated conditions (2 to 5 mM) needed to be maintained, semi-continuous, automated acetate
dosing was implemented using a computer feedback-controlled peristaltic pump with pre-
defined time interval or the anolyte Eh level as the reference set point in the LabVIEW™
feedback control program.
Chapter 3: Affinity of MFC Biofilm for the Anodic Potential
- 58 -
3.2.2 Calculation and Analysis
3.2.2.1 Determination of Voltage, Current and Power Generation
Fuel cell power output (P) was calculated from the voltage (V) measured across a known
resistor (adjusted by using a variable resistor box) and the current (I). An 8-channel PC parallel
relay board connected to a pre-defined array of resistors was controlled by the computer to allow
computer adjusted external resistance of the MFC. I was calculated according to Ohm‘s Law, I =
V/R, where V is the measured voltage and R (Ω) is the external resistance. Power output was
calculated according to: P=V x I. Since it was impractical to determine the surface area of the
granular graphite electrode used in this study and a significant part of the surface area of the
anode was inaccessible to bacteria, volumetric current density and power density were
determined by normalizing the values to the void volume of the anodic chamber (Rabaey and
Verstraete 2005). Current (in milliamperes, mA) was converted to electrons recovered by using
the following conversions: 1 Coulomb (C) = 1 Amp x 1 second, 1 C = 6.24 x 1018
electrons, and
1 mol = 6.02 x 1023
electrons (96,485 C/mol). Coulombic efficiency was calculated according to
Logan et al. 2006 (Logan et al. 2006).
3.2.2.2 Measuring the Effect of Anodic Potential on Microbial Activity by varying
External Resistance
The relationship between anodic potential and microbial activity in the MFC was
established by varying the external resistance of the MFC to attain steady potential of the anode.
This was achieved with three different approaches:
(1) Manual step changes of external resistance. This technique is commonly used in
many other studies mainly for polarization curve analysis (Logan et al. 2006);
(2) Constant sweeping of external resistance. A computer controllable relay board
equipped with a series of resistors was used to give a desired resistance output. As
such, the external resistance of the MFC could be systematically controlled at a fixed
sweeping rate (unit: Ω min-1
); and
(3) Switching at steady state the external resistance from a low level (e.g. 1 Ω) to open
circuit and observing the decrease of anodic potential. This approach examines the
bacterial capacity to charge up the anode under different conditions (e.g. substrate
saturated or limited conditions).
Chapter 3: Affinity of MFC Biofilm for the Anodic Potential
- 59 -
3.3 Results and Discussion
3.3.1 Characteristics of the MFC
A microbial fuel cell was set up as described above and operated for over 200 days.
During this time the current generated by the bacteria increased steadily. It was established that
the electron flow from anode to cathode was proportional to the rate of acetate oxidation by the
bacteria, and as soon as acetate in the anolyte was depleted, the electron flow in the cell dropped
to close to zero (Figure 3.1). A majority of electrons obtained from acetate oxidation was
retrieved as measured electron flow through the resistor (coulombic efficiency of about 83%)
(Figure 3.2).
Figure 3.1 Effect of external resistance (1 Ω and 1 MΩ) on acetate degradation ( ) and current
production (solid line) in the acetate-fed microbial fuel cell.
3.3.2 Initial Changes of Resistors result in Steady States of Different Microbial Activities
In order to test the effect of the anodic potential on the microbial acetate oxidation rate
the MFC was operated with different resistors under acetate saturated conditions (Figure 3.3).
As expected, a large external resistance (e.g. 160 Ω) did not allow a ready flow of electrons from
the anode to cathode and hence the anode stayed ―charged up‖ by the bacteria resulting in a low
anodic potential of less than -530 mV (vs. Ag/AgCl). For each resistance chosen, a new steady
state anodic potential established. This steady state is the result of the rate of microbial electron
Time (min)
Curr
ent (m
A)
Aceta
te C
oncentra
tion (m
M)
0
20
40
60
80
100
120
140
0 200 400 600 800
0
2
4
6
8
10
0
20
40
60
80
100
120
140
0 200 400 600 800
0
2
4
6
8
10
1 MΩ 1Ω 1 MΩ 1Ω
Acetate addition
Chapter 3: Affinity of MFC Biofilm for the Anodic Potential
- 60 -
delivery towards the anode being identical to the electron flow away from the anode to the
cathode and hence the calculated current can be used as a direct indicator of microbial acetate
oxidation rate (Figure 3.3D).
Figure 3.2 Current ( ) and Coulombic efficiency ( ) as a function of initial acetate
concentration in the acetate-fed microbial fuel cell. (Data were obtained after 16 days of
operation at 1 Ω.)
In general, the use of smaller resistors caused an increase in the anodic potential (Figure
3.3A and B). As expected, the microbial acetate oxidation rate increased with the anodic
potential as this not only provides a higher free energy change of the microbially catalyzed
reaction, but it also stimulates the re-oxidation of any electron shuttles (mediators) that may be
present in the system hence facilitating the electron transfer from bacteria to such mediators.
3.3.3 At a Certain Potential Range the Dependency of Microbial Activity on Anodic
Potential is Less Defined
More careful exploration of the relationship between microbial activity and the anodic
potential revealed that once an anodic potential of about -420mV (vs. Ag/AgCl) was reached, no
clear steady state values of current and potential could be obtained. Small decreases in resistance
did no longer increase the current but instead the anodic potential drifted rapidly to becoming
more positive (Figure 3.3B, at ~110 min).
Acetate Concentration (mM)
Curr
ent (m
A)
0
20
40
60
80
100
120
140
0.0 0.5 1.0 1.5
0
20
40
60
80
100
Ks = 0.36 mM
Cou
lom
bic
Effic
iency (%
) Vmax. = 140 mA
Chapter 3: Affinity of MFC Biofilm for the Anodic Potential
- 61 -
Figure 3.3 Changes in (A) external resistance; (B) anodic potential; (C) current and (D) acetate
oxidation rate in the acetate-saturated microbial fuel cell during a manual step-change of
external resistance experiment. (Note: the values of acetate oxidation rate were
calculated from the current after it had reached steady state. A new resistor was only
chosen after a new equilibrium between anodic potential and current had established.)
-550
-450
-350
-250
-150
-50
0
10
20
30
40
50
60
70
80
0
20
40
60
80
100
120
140
160
180
0
20
40
60
80
20 30 40 50 60 70 80 90
0
2
4
6
8
10
50 100 150 200 250 300
-540-520-500-480-460
0 10 20 30 40 50 60 70 80 90
05
101520
20 25 30 35 40 45 50
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 50 100 150 200 250 300
Time (min)
Aceta
te O
xid
ation R
ate
(mM
∙h-1)
Curr
ent (m
A)
Anodic
Pote
ntial
(mV
vs. A
g/A
gC
l)
Exte
rnal R
esis
tance
(Ohm
)
(B)
(A)
(D)
(C)
Chapter 3: Affinity of MFC Biofilm for the Anodic Potential
- 62 -
The unexpected drift in anodic potential was studied by allowing longer times for the
system to equilibrate (Figure 3.4). After initial steady values of anodic potential and current for
each resistor chosen, the potential started drifting at about -420 mV (vs. Ag/AgCl) when the
external resistance was decreased from 8 to 7 Ω (~8 min). The anodic potential was allowed to
drift resulting in a much higher quasi-steady state of about -165 mV (vs. Ag/AgCl) (~25 min).
Even by using slightly larger resistors the trend could not be reversed. Many repeat experiments
verified this to be a reproducible phenomenon. Overall, reproducible values of microbial activity
for anodic potentials between -350 and -420 mV (vs. Ag/AgCl) could not be obtained.
Figure 3.4 Changes of anodic potential (grey curve) and current (black curve) in the acetate-
saturated microbial fuel cell during downward and upward step-changes of external
resistance. (Dotted arrow shows the drifting of anodic potential after the external
resistance was decreased from 8 to 7 Ω at about 8 min.)
3.3.4 Apparent Maximum of Microbial Activity at Relatively Low Anodic Potentials
In spite of the unstable behavior of current and anodic potential between -350 and -420
mV (vs. Ag/AgCl) a maximum of microbial activity could be observed from a plot of acetate
oxidation rate versus anodic potential (Figure 3.5). Similar trends were observed at different
developmental stages of the anode biofilm, indicating that such phenomenon was reproducible
and independent of biomass concentration and composition in the MFC. The observed small but
marked maximum of apparent microbial activity at a rather low anodic potential between -350
and -420 mV (vs. Ag/AgCl) would indicate that the bacteria preferred a particular anodic
-450
-400
-350
-300
-250
-200
-150
-100
0 10 20 30 40 50 60 70 80
0
10
20
30
40
50
60
70
80
90
100
Time (min)
An
od
ic P
ote
ntia
l
(mV
vs. A
g/A
gC
l)
Cu
rren
t (mA
)
Exte
rnal R
esis
tan
ce
(Oh
m)
30
25
20
15
10
5
0
Current Anodic
Potential
External
Resistance 10Ω
9Ω 8Ω
7Ω 8Ω
10Ω
20Ω
30Ω
Chapter 3: Affinity of MFC Biofilm for the Anodic Potential
- 63 -
potential rather than being most active at the most positive anodic potential. This finding is
surprising for two reasons:
1. An increased anodic potential offers more free energy for the microbial reaction and
would be expected to stimulate microbial activity.
2. A lowering of resistance would normally increase the electron flow of MFC even if the
anodic reaction was not stimulated.
The fact that a lowering of resistance in a MFC caused a lower current is contrary to
general observations reported in the literature (Gil et al. 2003; Logan et al. 2006; Menicucci et al.
2006).
Figure 3.5 Dependency of acetate oxidation rate on the anodic potential of the acetate-saturated
microbial fuel cell. MFC after 5 ( ) and14 ( ) days of operation and after long term
operation with lower currents (with oxygen cathode) ( ).
3.3.5 An Activity Maximum is only obtained when moving from Low to Higher Anodic
Potentials
The phenomenon that at an anodic potential of about -420 mV (vs. Ag/AgCl) the
bacterial activity appeared to have a maximum, was observed for a number of different
conditions as long as incrementally smaller values of resistance were chosen (i.e. sweeping the
potential from negative values towards zero). However, when using incrementally larger rather
0.0
0.5
1.0
1.5
2.0
2.5
3.0
-550 -450 -350 -250 -150 -50 50
Anodic Potential (mV vs. Ag/AgCl)
Aceta
te O
xid
ation R
ate
(m
M h
-1)
Chapter 3: Affinity of MFC Biofilm for the Anodic Potential
- 64 -
than smaller resistors to document the current-anodic potential relationship, the previously
observed current maximum at around -420 mV (vs. Ag/AgCl) was not found (Figure 3.6).
Specific sweeps of resistance changes both upwards and downwards (sweeping rate of 0.15 Ω
min-1
) established clearly that the observed peak in current was not due to a true maximum in
microbial activity or fuel cell performance but due to a ―hysteresis effect‖ caused only by
increasing the anodic potential. By forming an average of the measurements taken from different
experiments using both increasing and decreasing resistor changes, a reasonably reliable
relationship between bacterial activity and anodic potential could be obtained (Figure 3.6).
Figure 3.6 MFC current ( = by decreasing external resistance; = by increasing external
resistance; = average current) as a function of anodic potential of the acetate-saturated
microbial fuel cell in a constant external resistance sweeping experiment. (Sweeping rate
~0.15 Ω min-1
, ranged from 1 – 78 ohm.)
3.3.6 Open Circuit Drop in Anodic Potential
Another way of investigating the bacterial capacity to charge up the anode is by
recording the build-up of anodic potential at open circuit with a known bacterial acetate
oxidation rate (and hence microbial electron flow generation) (Figure 3.7). After switching the
external resistance from 1 ohm to open circuit, the anodic potential dropped to about -420 mV
(vs. Ag/AgCl) (Figure 3.7A). Then, exactly in the area of the instability of anodic potential and
current as described above, the anodic potential stayed almost constant at about -420 mV (vs.
Ag/AgCl) for about 18 minutes (Figure 3.7A) before droping at a higher rate down to its final
Anodic Potential (mV vs. Ag/AgCl)
Cu
rre
nt
(mA
)
0
10
20
30
40
50
60
70
80
-500 -450 -400 -350 -300 -250 -200 -150 -100 -50
Chapter 3: Affinity of MFC Biofilm for the Anodic Potential
- 65 -
level at about -480 mV (vs. Ag/AgCl). The same trend could be observed for bacteria that were
starved of acetate. The anode biofilm in Figure 3.7B was starved at high anodic potential for
over 24 hours to oxidize not only all acetate but also any internal storage polymers. This leaves
maintenance respiration as the only electron donating reaction for current production. This
tendency of a decelerating drop in anodic potential between -415 and -425 mV (vs. Ag/AgCl)
suggests one of two possibilities:
1. The bacterial electron release was now ―buffered‖ by the presence of an oxidized
electron mediator (capacitance). In this case, instead of decreasing the potential further
the mediator is being reduced (charged up).
2. Alternatively the slower drop in potential could be due to the bacteria becoming less
active.
Figure 3.7 Profile of anodic potential (black curve) and anolyte Eh (grey curve) of a microbial
fuel cell under (A) acetated-saturated and (B) acetated-depleted conditions during open
circuit. (Dashed line indicates the anodic potential at which quasi-steady states were
attained for a short period before dropping further to the lowest steady level.)
3.3.7 Detailed Interpretation of the Dependency of Microbial Activity on Anodic Potential
Figure 3.8 illustrates a plot of microbial activity and MFC power output as a function of
terminal electron acceptor availability (anodic potential). The curve was characterized by the
following values:
1. At anodic potentials lower than about -420 mV (vs. Ag/AgCl) the microbial activity was
limited by the anodic potential or by the availability of sufficient oxidized electron
mediator that is used for the transfer of electrons to the anode.
-500
-400
-300
-200
-100
0
100
0 10 20 30 40 50 60 70 80 90 0 60 120 180 240 300 360 420 480Time (min)
Po
ten
tia
l (m
V v
s. A
g/A
gC
l)
1Ω 1Ω
Anode
Open Circuit
(A) Substrate Saturated (B) Substrate Depleted
Anolyte
Anode
Anolyte
Open Circuit
Time (min)
Chapter 3: Affinity of MFC Biofilm for the Anodic Potential
- 66 -
2. There was no significant microbial activity at anodic potentials less than -480 mV (vs.
Ag/AgCl). This is most likely caused by the reducing capacity of the electron donor
acetate with standard redox potential of -480 mV (vs. Ag/AgCl). The fact that some
current was observed at anodic potentials lower than that could be explained by the
product and substrate concentrations affecting the actual Eh relative to the standard Eh for
standard conditions. Moreover, bacteria may have the capacity to generate more powerful
reducing agents in the process of assimilation.
3. There was a Plateau of microbial activity for anodic potentials between -200 and about -
420 mV (vs. Ag/AgCl) representing excess availability (saturation) of electron acceptors
(oxidized mediators or anode). The microbial metabolism was not limited by the capacity
of the anode to serve as electron sink but by other metabolic bottlenecks such as acetate
uptake, TCA cycle or similar.
4. In biochemical terms a half saturation constant for anodic potential could be estimated. In
analogy to the Michaelis-Menten half saturation constant (here termed kAP value) it would
represent the anodic potential (or the concentration of oxidized mediator) at which the
metabolic rate (here measured as current) reaches half its maximum rate. This kAP value
was about -455 mV (vs. Ag/AgCl) for our acetate driven MFC and was independent of
the oxidation capacity of the cathodic half cell.
5. There was a critical anodic potential at about -420 mV (vs. Ag/AgCl) (APcrit., Figure 3.8)
beyond which further increased anodic potentials no longer stimulated bacterial activity
significantly, showing that now the biochemical electron flow was saturated by the
availability of the terminal electron acceptor (in this case the oxidized mediators for the
anode or the anode itself).
The significance of the above described critical anodic potential (APcrit.) is that it
characterizes the power output of the MFC (Figure 3.8). At anodic potentials below the APcrit.
the power output of the fuel cell increases with lower resistors providing a higher anodic
potential and enabling a greater microbial electron flow. At anodic potentials higher than APcrit.
the bacteria cannot generate significantly more electron flow as they are saturated by the
availability of suitable electron acceptors. Here, a further increase in anodic potential does not
yield more power because the current no longer increases substantially while the voltage of the
MFC diminishes due to increasing anodic potential (Figure 3.8). This suggests that the maximal
Chapter 3: Affinity of MFC Biofilm for the Anodic Potential
- 67 -
power output of a MFC is close to the point where the bacteria reach saturation of their electron
accepting mechanism by the anode.
Figure 3.8 Percentages of maximal microbial activity ( = manual R step-change; = constant
R sweeping; solid curve) and MFC power output ( = manual R step-change; =
constant R sweeping; dashed curve) as a function of anodic potential of an acetate-
saturated microbial fuel cell. (kAP = half saturation constant for anodic potential, ~ -455
mV vs. Ag/AgCl; APcrit. = critical anodic potential, ~ -420 mV vs. Ag/AgCl); Notes:
constant R sweeping rate was ~0.15 Ω min-1
; catholyte Eh ranged from 225 to 232, and
260 to 300 mV (vs. AgAgCl) for constant R sweeping and manual R step change
experiments, respectively.
The observation that a half saturation value can be determined and is in analogy to the
half saturation constants for the concentration of soluble substrates seems quite interesting and is
not published yet. It is expected to be useful for modeling and optimization of microbial fuel
cells and the relative comparison of different microbial consortia at the anode. However, as it is
known for kinetic values obtained for activated sludge (e.g. the kO2, the half saturation
concentration for the terminal electron acceptor oxygen) the APcrit. and kAP values are not
expected to be discrete values for all other MFC systems. However the general behavior is
widely used for modeling and process optimization. The discrete value would depend on the
thickness of the biofilm, the mass transfer limitation in the anodic chamber, reactor
configuration, the species composition present, similar to the kO2 value for activated sludge
bacteria.
0
20
40
60
80
100
-500 -400 -300 -200 -100 0
A Pcrit.
K AP
Anodic Potential (mV vs. Ag/AgCl)
Perc
enta
ge o
f M
axim
al A
ctivity
and P
ow
er
(%)
Chapter 3: Affinity of MFC Biofilm for the Anodic Potential
- 68 -
3.3.8 Redox Capacitance as an Explanation of Apparent Current Maximum
At this stage we assume that the peak in current observed during stepwise increases of
the anodic potential (but not during stepwise decreases) was not due to increased bacterial
activity but to reducing power that had been stored at the previous, lower potentials. This
tendency was also indicated by the Eh trends in the anodic half cell (Figure 3.7). It is not likely
that the well known phenomenon of bacterial storage (e.g. as PHB) formation (under anaerobic
conditions) and its oxidation under aerobic condition is responsible for this reproducible
observation because: a) the time intervals were only short, b) the same phenomenon could be
observed with acetate starving and acetate saturated cells. The fact that when using non-
equilibrium conditions by moving stepwise down with the resistors gave a different curve than
by moving stepwise up is in analogy to cyclic voltametry. Even though the ―sweeping rate‖ used
in our experiments is very low in comparison the resultant lines suggest that capacitance has
build up in the system.
However, it goes beyond the scope of this study to investigate the source (mediators in
the bulk solution, inside the biofilm or as part of the electrode) capacitance of this system.
However it can be stated that significant charge storage is observed. If the capacitance was due
to microbial shuttling compounds, then an approximate mid-point potential (or standard
potential) between -350 and -450 mV (vs. Ag/AgCl) would be indicated (Figure 3.6, 3.7).
3.3.9 Implication of Findings on the Design and Operation of MFCs
The fact that the microbial activity in MFC obeys the same kinetic type for the anode
than described for soluble species (Michaelis-Menten kinetics) is expected to be helpful in
evaluating different biofilms, bacteria species or environmental conditions. Similar to the
saturation kinetics of microbial metabolism of soluble species, it is expected that some biofilm
populations have a high affinity for the anode while others display a high maximum activity.
With the new insight about the relationship between microbial activity and anode
potential, an optimal operating condition of MFC could be defined by controlling or maintaining
the anodic potential at the APcrit. level at which the maximal bacterial activity and power output
of the MFC could be attained. Further, knowing the actual affinity for the anode and maximum
microbial activity in a MFC is expected to help with diagnosing fuel cell malfunctions.
- 69 -
4 A New Approach for in-situ Cyclic Voltammetry of a
Microbial Fuel Cell Biofilm without using a
Potentiostat
(This chapter has been published in Bioelectrochemistry, (2009) 74: 227-231)
Chapter Summary
Electrochemically active bacteria in a microbial fuel cell (MFC) usually exist as a
biofilm attached to an electrode surface. Conventional cyclic voltammetry using potentiostat is
considered as a powerful and reliable method to study electrochemical behavior of MFC biofilm.
In this chapter, a new approach to evaluate redox behavior of an electro-active MFC biofilm
without using a potentiostat was proposed and validated. Analogous to a conventional cyclic
voltammetry study, the biofilm-electrode potential was controlled by computer-feedback
controlling the external resistance of an operating MFC. In this way, the MFC can still operate
as a ―fuel cell‖ without being ―interrupted‖ by an external device (i.e. potentiostat) that normally
does not belong to the system.
Relationship between current and biofilm-electrode potential was obtained and showed
agreement with a potentiostat-controlled method under similar experimental conditions. The
method could be added to our technical repertoire for analysis of bacterial mediator involved in
the exocellular electron transfer of a MFC-biofilm, and it could potentially serve as a practical
process monitoring method for MFC operation. The application of computer-control
components should be further explored to facilitate control, diagnosis as well as optimization of
MFC processes.
Chapter 4: A New Approach for CV analysis without using a Potentiostat
- 70 -
4.1 Introduction
In light of the increasing demand of renewable and sustainable energy, microbial fuel
cell (MFC) technology becomes one of the most rapidly developing environmental
biotechnologies (Logan and Regan 2006b; Lovley 2006b; Rabaey and Verstraete 2005). Yet, the
technology is still in its infancy and needs further research to improve understanding and
performance of the process. MFC is a bacteria-catalyzed electrochemical device, in which a
bacterial biofilm on the anode transfers electrons from the oxidation of an organic substrate to
the anode enabling a current flow from anode to cathode. The presence of a suitable resistive
load in the external circuit enables the production of electrical power.
Arguably the most interesting phenomenon in MFC electron flow is the transfer of
electrons from the bacteria to the anode surface. Although different mechanisms have been
proposed, the most widely accepted view is that electron mediators are produced by the bacteria
for this transfer. Bacterial mediators (also termed electron shuttles or electron carriers) have also
been found in other microbial reactions involving an insoluble solid electron acceptor such as
ferric iron or elemental sulfur (Lovley et al. 2004; Stams et al. 2006). After the bacteria have
reduced an oxidized mediator as their terminal electron acceptor, it will then ―shuttle‖ the
electrons to the electron accepting anode where it becomes reoxidized, and becomes available to
the bacteria again (Rabaey et al. 2004; Rabaey et al. 2007; Schroder 2007).
The redox property of the bacterial mediators can be seen as critical for the effective
operation of the MFC. In the literature, redox properties of mediators were commonly studied by
using cyclic voltammetry (CV) method. This method is considered as a standard technique for
studying redox behavior of solutions (Bard and Faulkner 2001; Logan et al. 2006; Rabaey et al.
2005a; Rabaey et al. 2004; Scholz 2002; Sund et al. 2007).
In CV, the potential of a working electrode is scanned forward and backward at a
predefined rate (commonly ranges from 5 to 100 mV·s-1
). The presence of any mediators is
detected from the fact that a current is produced during the cyclic oxidation and reduction of the
mediator, provided that the mediator can accept (/ donate) electrons from (/ to) the working
electrode. The mid potential of the mediator can then be obtained from the current-voltage plot
(i.e. cyclic voltammogram) (Bard and Faulkner 2001; Marken et al. 2002).
Using conventional CV (i.e. using potentiostat to control a half cell potential) to evaluate
electrochemical behavior of an electrode-associated biofilm has several advantages. For
Chapter 4: A New Approach for CV analysis without using a Potentiostat
- 71 -
examples: (i) the target electrode processes (e.g. biofilm-anode) can be studied separate from the
opposing electrode (i.e. cathode) reaction and separate from other side effects such as ion
transfer processes across the separating membrane; (ii) the potential can be varied in a wide
range, even exceeding the potential range of a chemical cathode.
However, in conventional CV the MFC has to be disconnected from the external power
user (i.e. external resistance) in order to connect with a potentiostat. From a practical viewpoint,
this may lead to an operation down time of the process during which the MFC can no longer
generate power. In this chapter, a new method of controlling the MFC external resistance to
perform cyclic voltammetric analysis of an anode biofilm is proposed. In this way, the MFC can
still operate as a ―fuel cell‖ without being ―interrupted‖ by an external device (i.e. potentiostat)
that normally does not belong to the system. Further, the optimal external resistance of a MFC at
which the MFC delivers the highest power may also be defined or maintained by using the
proposed method.
It is the aim of this study to develop a practical approach of performing CV analysis for
MFC process without using a potentiostat. This method will supplement the well established
conventional CV method of evaluating redox properties of an electrode-associated biofilm. It is
hypothesized that CV of a MFC biofilm could be done by feedback-controlling the external
resistance of an electricity producing MFC without the need of a potentiostat. This means no
additional equipment is needed other than the computer-controlled MFC itself. The MFC was
set-up and operated as described in Chapter 3.
Chapter 4: A New Approach for CV analysis without using a Potentiostat
- 72 -
4.2 Experimental Section
4.2.1 Underlying Principle of the Proposed Method
Under close-circuit operation of an active MFC (i.e. electrons are allowed to flow readily
from the biofilm anode to the cathode), there is a continuous reduction and reoxidation of the
mediator involved in the electron transfer from the bacteria and the anode. The reduction
reaction can be realized by using a larger external resistance, resulting in an accumulation of
reduced mediator and hence charging up the biofilm-anode (i.e. decreased electrode potential).
While the oxidation reaction is facilitated by using a smaller external resistance, resulting in
faster anode discharge (i.e. increased electrode potential).
Similar to the theory of traditional CV our proposed method aims to recognize mediators
by their capacity to donate or accept electrons when the electrode potential is continuously
shifted (scanned). However, instead of using an external reducing power source (i.e.
potentiostat), our proposed method relies solely on the bacteria to reduce their self-produced
mediators using electrons obtained from their substrate oxidation. In this case the electrode
serves only as the final electron acceptor. This method can selectively detect redox species
involved exclusively in the bacterial electron transfer chain, and therefore it has the potential to
offer more specific information compared to traditional CV in evaluating redox properties of a
MFC biofilm.
4.2.2 Microbial Fuel Cell
A two-chamber MFC made of transparent Perspex was used in the present study. Figure
4.1 shows the schematic of the reactor and its peripheral settings. The two chambers were of
equal volume and dimension (316 mL (14 cm 12 cm 1.88 cm)). They were physically
separated by a cation-selective membrane (CMI-7000, Membrane International Inc.) with a
surface area of 168 cm2. Conductive graphite granules (El Carb 100, Graphite Sales, Inc.,
Chagrin Falls, OH, U.S.A., granules 2-6 mm diameter) were used as both the anode and cathode
electrodes, which decreased the internal liquid volume from 316 to 120 mL. This internal void
anolyte volume was used to calculate the volumetric current and power densities of the MFC.
The biofilm in the anode chamber was originated from an activated sludge collected
from a local wastewater treatment plant. The growth medium (anolyte) composed of (mg L-1
):
Chapter 4: A New Approach for CV analysis without using a Potentiostat
- 73 -
NH4Cl 125, NaHCO3 125, MgSO4·7H2O 51, CaCl2·2H2O 300, FeSO4·7H2O 6.25, and 1.25 mL
L-1
of trace element solution, which contained (g L-1
): ethylene-diamine tetra-acetic acid (EDTA)
15, ZnSO4·7H2O 0.43, CoCl2·6H2O 0.24, MnCl2·4H2O 0.99, CuSO4·5H2O 0.25, NaMoO4·2H2O
0.22, NiCl2·6H2O 0.19, NaSeO4·10H2O 0.21, H3BO4 0.014, and NaWO4·2H2O 0.050. The
medium was complemented with 50 mM phosphate buffer to maintain a constant pH of 7. Yeast
extract solution was periodically added (ca. every 3 to 5 days) to the medium (50 mg L-1
final
concentration) as bacterial growth supplement. Air-saturated solution of potassium ferricyanide
(100 mM), K3Fe(CN)6 (Sigma-Aldrich, Inc., purity ca. 99%) complemented with 100 mM
phosphate buffer was used as the catholyte. Detail operation and monitoring of the MFC during
the initial start-up period were described in our previous paper (Cheng et al. 2008).
Figure 4.1 Schematic of the computer-feedback controlled microbial fuel cell used in the
present study.
4.2.3 Cyclic Voltammetry by Feedback-Controlling the External Resistance
The biofilm-electrode potential was controlled by selecting appropriate external
resistance of the MFC. In initial experiments, the external resistance of the MFC was
continuously changed in a linear fashion by using an eight channel computer controlled relay
board (Ocean Controls, Victoria, Australia; www.oceancontrols.com.au, Photo 4.1) (Figure
DAQ Board
Computer
Control/
Monitoring by
LabVIEW™
Stirrer
1M NaOH
Acetate Feed
Moisten
Air
Variable External Resistor
(PC-Controllable Relay Board)
Anolyte pH Anolyte Eh
Catholyte Eh
Ag/AgCl
Reference
Electrode
Biofilm-Anode
(Granular Graphite)
Counter Electrode
(Granular Graphite)
Cation Exchange Membrane
e- e-
Recirculation
Pump
MF
C V
olta
ge
Bio
film
-Ele
ctr
ode
Po
ten
tial
Catholyte
Graphite Rod
( Diameter =
5mm )
Out In
DAQ Board
Computer
Control/
Monitoring by
LabVIEW™
Stirrer
1M NaOH
Acetate Feed
Moisten
Air
Variable External Resistor
(PC-Controllable Relay Board)
Anolyte pH Anolyte Eh
Catholyte Eh
Ag/AgCl
Reference
Electrode
Biofilm-Anode
(Granular Graphite)
Counter Electrode
(Granular Graphite)
Cation Exchange Membrane
e- e-
Recirculation
Pump
MF
C V
olta
ge
Bio
film
-Ele
ctr
ode
Po
ten
tial
Catholyte
Graphite Rod
( Diameter =
5mm )
Out In
Chapter 4: A New Approach for CV analysis without using a Potentiostat
- 74 -
4.2A). The relay was directly controlled by the parallel port output (8 bits) from the computer.
The slow rate linear sweeping of external resistance (0.69 min-1
) had reproducibly yielded the
so-called polarization and power density curves which are commonly used to assess MFC
performance (Logan et al. 2006) (Figure 4.2B). Furthermore, this constant resistance sweeping
technique was validated in our previous study to establish the relationship between the current
and the anodic potential (Figure 4.2C), from which the affinity of an electrochemically active
biofilm for the electrode potential could be determined (Chapter 3, Cheng et al. 2008).
Photo 4.1 Computer controllable relay board served as the variable resistor of the MFC.
In practice, the external resistance can be controlled in a number of ways, including the
use of an electronically adjustable resistor. In our experiments the switching of 8 individual
resistors via the relay board allowed up to 256 different resistance values, which was sufficient
for adequate control of the biofilm-electrode potential. This improved way of changing the
resistors allowed a relatively linear scan of potential over time (Figure 4.3A and B). This was
accomplished as follows: a computer program was designed using LabVIEW™ programming
software (Version 7.1) to allow the operation of the MFC at a constant electrode (here anode)
potential setpoint. To maintain the electrode potential setpoint the external resistance was
controlled (see above) by a simple feedback loop in the program, which enabled not only the
operation to a desired potential setpoint but also the programmed stepwise shift of the potential.
This could allow a linear potential change over time (here 0.1 mV·s-1
), as it is known in CV (i.e.
upward and downward scans represent shifts of potential toward a less and a more negative
values, respectively) (Figure 4.3B).
Chapter 4: A New Approach for CV analysis without using a Potentiostat
- 75 -
Figure 4.2 Constant sweeping of the external resistance of the computer-feedback controlled
acetate-saturated microbial fuel cell (0.69 min-1
). (A) Resistance vs. time; (B)
Polarization and power density curves; Note: ↑ or ↓ R represents increasing or decreasing
sweep of external resistance, respectively. (C) Current vs. electrode potential.
0
20
40
60
80
-600 -500 -400 -300 -200 -100 0
Electrode Potential (mV vs. Ag/AgCl)
Curr
ent
(mA
)
0
100
200
300
0 5 10 15Time (h)
Resis
tance (
Ohm
) .
0
100
200
300
400
500
600
700
0 200 400 600
Current Density (A m-3
)
Voltage (
mV
)
0
50
100
150
200
250
300
Pow
er D
ensity
(W m
-3)
Voltage: ↑ R
Voltage: ↓ RPower: ↑ R
Power: ↓ R
E/ mV (vs. Ag,AgCl (3M KCl))
Voltage/ m
V
Curr
ent/
mA
Power D
ensity
/ W∙m
-3
A.
B.
C.
Time/ h
Resis
tance/ O
hm
Current Density/ A∙m-3
Chapter 4: A New Approach for CV analysis without using a Potentiostat
- 76 -
It should be noted that these experiments were carried out with a fully developed biofilm
that was saturated with organic substrate supply (here acetate, about 10 mM) which served as a
relatively constant source of reducing power in place of a potentiostat. The ultra low scan rate of
0.1 mV·s-1
was selected to allow sufficient time for the biofilm (microbially-reduced mediator)
to interact with the oxidizing electrode. As a result, the method enabled the detection of
mediators that were not only located within the boundary layer of the electrode-solution
interface, but also mediators far away from the electrode (i.e. confined within the biofilm or
inside the bacterial cells). Using this method with scan rates that are normally used in traditional
CV, no significant peaks/ valleys were detected (data not shown).
4.2.4 Cyclic Voltammetry using a Three-electrode Potentiostat
In the present study, the new method was validated by comparing with results obtained
from a potentiostat-controlled method. In this Control experiment, the potential of the same
biofilm-electrode was regulated by using a three-electrode scanning potentiostat (Model no.362,
EG&G, Princeton Applied Research, Instruments Pty. Ltd.). The working, counter and reference
electrodes of the potentiostat were connected to the anode, cathode and the silver-silver chloride
reference electrode of the acetate-saturated MFC, respectively. The electrode potential was
changed at a similar scan rate of 0.1 mV·s-1
to allow reasonable comparison with the new
method. To assure quality of the results, all experimental data (obtained with both the new
method and the Control method) were validated for reproducibility by running for at least three
repetitive cycles in each experimental testing.
Chapter 4: A New Approach for CV analysis without using a Potentiostat
- 77 -
4.3 Results and Discussion
4.3.1 Feedback Controlling the External Resistance of a MFC enables Cyclic
Voltammetry of MFC Biofilm
The current obtained during the scans of biofilm-electrode potential was recorded and
showed characteristic and reproducible patterns (Figure 4.3C). In theory, decreasing electrode
potentials will thermodynamically impede the electron flow from the bacteria to the electrode
and hence resulting in lower currents. However, depending on whether the potential was
gradually increased or decreased, different current lines were obtained (Figure 4.3B and C). This
is indicative of charging and discharging of bio-electrochemically active compounds. To
visualize more clearly the redox behavior of the electrochemically active biofilm the current was
plotted as a function of electrode potential (i.e. cyclic voltammogram).
Figure 4.3 Time profiles of (a) external resistance; (b) electrode potential; and (c) current during
the computer-feedback control of the external resistance of an acetate-saturated microbial
fuel cell. Electrode potential scan rate was 0.1 mV·s-1
.
-450
-350
-250
-150
Ele
ctr
od
e P
ote
ntia
l
(mV
vs. A
g/A
gC
l)
0
20
40
60
80
100
0 1 2 3 4
Cu
rre
nt (m
A)
0
20
40
60
80
Re
sis
tan
ce
(O
hm
)
Time (h)
A.
B.
C.
Chapter 4: A New Approach for CV analysis without using a Potentiostat
- 78 -
To test the hypothesis of the present study, the new method was verified by comparing
with results obtained from a potentiostat-controlled method. Figure 4.4 indicates that both
computer-feedback- and potentiostat-controlled experiments revealed similar results: the upward
shift of electrode potential from about -380 to -200 mV vs Ag/AgCl gave significantly higher
currents compared to currents obtained from the downward shift in the same potential range.
While at potentials outside this range, the resultant currents recorded from both upward and
downward scans were virtually identical. Many repeat experiments verified this to be a
reproducible phenomenon. These reproducible equilibrium currents (i.e. bacterial activity) were
independent on the direction of potential scans and indicated that bio-electrochemically active
species were not active (charging/ discharging) at these potentials (Figure 4.4).
Figure 4.4 Comparison between the current-potential plots (cyclic voltammograms) obtained by
(a) computer-feedback controlling the external resistance and (b) by using a potentiostat
in an acetate-saturated microbial fuel cell. Scan rates of the electrode potential in both
methods were 0.1 mV∙s-1
; dotted arrows indicate the direction of the potential scan; the
difference between the currents obtained from the two methods was due to the different
times of the two experiments.
The drop in current when the potential decreased from about -200 to -330 mV vs
Ag/AgCl suggests one of two possibilities:
0
20
40
60
-450 -350 -250 -150
Electrode Potential (mV vs. Ag/AgCl)
Cu
rre
nt
(mA
)
Feedback Controlling of Resistor
Potentiostat-Controlled
35
40
45
-450 -350 -250 -150
E/ mV (Ag,AgCl (3M KCl))
Cu
rre
nt/
mA
Chapter 4: A New Approach for CV analysis without using a Potentiostat
- 79 -
(1) Bacteria became less active at lowered electrode potentials;
(2) The electrons released by the bacteria were accepted and stored by a ―pool‖ of alternative
electron acceptor (i.e. oxidized mediators) instead of the electrode (capacitance from
mediator).
Since a further decrease in potential from -330 to -380 vs Ag/AgCl resulted in increasing
currents again, the first possibility seems unlikely (Figure 4.4). The observed current ―valley‖ in
both experiments was likely caused by the redox reactions of the mediator. This is supported by
the fact that during upwards shifts of potentials there was no ―valley‖. From the cyclic
voltammograms obtained in these modified CV experiments (Figure 4.4), one could obtain the
mid potential of the mediator from the potential at which the upward and downward curves are
most widely apart (i.e. highest capacitance). In our acetate fed MFC biofilm, such potential
value was about -330 mV vs Ag/AgCl.
It has to be noted that as MFC is a bio-electrochemical system, the anode associated
biofilm can grow and develop over time. Hence, over hours of operation during the CV
experiments, it is expected that the current output obtained at even the same electrode potentials
would change. This explains the difference between the current-potential curves in Figure 4.4.
Nevertheless, our results suggested that both the proposed and the conventional CV methods
indicated a same redox behavior of the same biofilm.
However, it is speculative to characterize the nature of a redox active species involved in
the anodic electron transfer in a MFC (i.e. whether it is purely membrane associated such as
cytochrom c or nanowire (i.e. direct electron transfer) or soluble (i.e. indirect electron transfer)
solely based on results obtained from both the proposed and the conventional CV techniques.
According to earlier findings (Bond and Lovley 2005; Rabaey et al. 2007), bacterial soluble
mediators appear to be confined mainly within the biofilm rather than in the bulk medium. This
property was due to an electrostatic interaction between the soluble mediator and the anode
(Rabaey et al. 2007). Therefore, CV of a biofilm-anode cannot be used as a stand-alone method
to ascertain the physiochemical nature of the mediator involved in the electron transfer process.
Overall, the new method has demonstrated a good controllability of the biofilm electrode
potential, and it showed a similar capacity to the conventional potentiostat-controlled method in
evaluating in situ electrochemical properties of a MFC biofilm. Further, provided that the
electron donor is not limiting in the system, evaluation of the electro-active biofilm in a MFC
Chapter 4: A New Approach for CV analysis without using a Potentiostat
- 80 -
can be conducted without the need of a potentiostat. To our knowledge, this is the first report in
the literature to control the biofilm-electrode potential by feedback controlling the external
resistance of a MFC in a way similar to a CV analysis. Redox information about the mediator(s)
involved in the bacterial electron transfer as well as the affinity of the electro-active biofilm on
the electrode potential (i.e. values of critical electrode potential and half-saturation electrode
potential as defined in (Chapter 3, Cheng et al. 2008; Marcus et al. 2007) could be obtained with
minimal disturbance of the bioprocess.
4.3.2 Limitations and Implications of the New Method
The established technique has two major limitations that are needed to be addressed in
future studies.
1. It requires the presence of bacterial substrate. In MFC, as the electrons flow exclusively in
one direction starting from the substrate bacteria bacterial mediator(s) anode
external resistor cathode terminal electron acceptor, sufficient substrate (electron
donor) must be provided to the biofilm during the analysis. While in the absence of
substrate, there is no reducing power to reduce the involved mediator(s) and hence the
electrode. Perhaps, such criterion could be further explored for investigating specific
electron transfer processes with different substrates.
2. It requires an efficient cathodic reaction. The cathodic reduction process should not be
limiting over the electrode potential range in which bioelectrocatalysis of the bacterial
mediators is taking place. In the present study, the use of ferricyanide solution enabled a
stable and strong cathodic reduction, allowing the effective control of anodic potential by
using the proposed method. However, it is expected that with the continuous improvement
of MFC cathodic process, such limitation could be alleviated.
Apart from the application as discussed above, the established technique may also be
useful to gain other important insights for MFC process control. For instance, selection of
external resistor is considered as critical for optimization of MFC process. Maximal and
sustainable power output can only be achieved when the MFC is running with an appropriate
external resistance. While most studies reported thus far have only arbitrarily selected the
external resistance to operate their MFCs, and very often the power outputs of their systems
were reported by using an inappropriate external resistor (Menicucci et al. 2006). By feedback
Chapter 4: A New Approach for CV analysis without using a Potentiostat
- 81 -
controlling the external resistance of the MFC using either current/ power generation as the
reference parameter, we could in theory operate the MFC at its optimal current/ power
production point (see supplementary figure in Appendix 3). Overall, it is expected that more
sophisticated and intelligent application of computer control components will be emerged in the
future to deal with optimization and diagnosis of such dynamic bio-electrochemical process.
- 82 -
5 An Anodophilic Biofilm prefers a Low Electrode
Potential for Optimal Anodic Electron Transfer: A
Voltammetric Study
Chapter Summary
The knowledge of electron transfer between anodophilic biofilms and their associated
electrodes is important for both the fundamental understanding and optimization of BES. In this
Chapter, a series of potentiostatic experiments were conducted to explore in details the electron
transfer behavior of an active anodophilic biofilm in a BES. The results suggest that using ultra
slow potential scan rates of 0.01mV s-1
(between -450 and -150 mV) an actively current
producing bacteria revealed peaks of current that could not be detected with traditional scan
rates or with starving bacteria. It appeared as if the bacterial electron transfer to the electrode
had an optimum potential at around -230 mV vs. Ag/AgCl. At another potential of -310 mV vs.
Ag/AgCl the presence of a redox species (or redox system) was indicated by the fact that a peak
during upwards (from -ve towards +ve) potential scanning was not found during downwards
scanning. The draining of the electrolyte and replacing by fresh solution did not affect the
electrochemical characteristics of the anodic biofilm, demonstrating that the electrochemically
active species were predominately present in the biofilm at the electrode rather than in the bulk
medium.
The fact that the presence of the redox peaks was dependent on the presence of the
bacterial electron donor (here acetate) suggested that the reduction of the mediators (redox links)
was of biochemical and not purely electrochemical nature. The fact that the complexity of the
bioelectrochemical behavior of an electrochemically active biofilm could only be revealed by
using very low scanning rates indicates that the redox process in the biofilm-anode system was
slow compared to abiotic mediator-electrode systems. Overall, the use of low scan rates for CV
analysis may be an improved way to characterize the electrochemical behavior of an anodophilic
biofilm in BES.
Chapter 5: An Anodophilic Biofilm prefers a Low Anodic Potential
- 83 -
5.1 Introduction
As mentioned in Chapter 1 and 3, the potential of the anode determines the amount of
Gibb free energy conserved by the anodophilic bacteria in a BES (denoted as ΔGo‘
bacteria in
Figure 1.5, page 14). From a thermodynamic viewpoint, it is expected that the anodophilic
bacteria could conserve more energy at a higher anodic potential and hence a more positive
anodic potential is expected to encourage establishment of electrochemically active bacteria on
the electrode surface. However, in order to maximize electricity generation in MFC or to
minimize the extra energy input in MEC (both are denoted as ΔGo‘
Elec in Figure 1.5), a minimal
amount of ΔGo‘
bacteria is preferred as it keeps the anode at a lower redox state and hence the
anode can be more reducing. Further, if the bacterial energy gain becomes too large (e.g. at a
very positive anodic potential), the electrical energy output will be diminished and electrons
from the substrate may be diverted toward an assimilation metabolism, leading to excessive
biomass yield but reduced electron recovery (Pham et al. 2009). This may lead to a question of
whether an anodophilic biofilm would prefer or adjust their metabolism to a certain anodic
potential to meet the antagonistic requirements between bioenergetics and thermodynamics of a
BES? In the literature, there are evidences suggesting that anodophilic biofilms exhibit
saturation behavior with electrode potentials (Chapter 3, Cheng et al. 2008; Torres et al. 2008a;
Torres et al. 2007). Yet, no clear evidence is available in the literature to elucidate the existence
of a preferred electrode potential for optimal anodic electron transfer. Moreover, the underlying
process of anodic electron transfer from electrochemically active microorganisms to their
associated electrode (bioanode) is highly complex (Fricke et al. 2008; Pham et al. 2009;
Schroder 2007).
Anodophilic bacteria in BES commonly exist as a biofilm attached to the electrode, the
diffusion or mass transport of soluble mediators (if any) either toward or away from the
electrode surface is limited as the biofilm may block the direct electrochemical contact between
the mediators in the bulk and the electrode. Therefore, to better qualify or quantify the redox
behavior of an electrochemically active biofilm using cyclic voltammetry, sufficient time is
required to allow the redox active species to response to the electrode potential.
In Chapter 4, a new approach of performing voltammetric analysis in a BES without
using an external electrochemical device (a potentiostat) has been described. However, as
mentioned in Section 4.3.2 this new method adheres to two inflexibilities: 1) the anodic biofilm
must be saturated with the electron donor substrates; and 2) a powerful cathodic reaction must
Chapter 5: An Anodophilic Biofilm prefers a Low Anodic Potential
- 84 -
be provided. While the use of a potentiostat to investigate the electron transfer of an
electrochemically active biofilm can avoid disturbance and artifact due to the second electrode
(i.e. cathode, if anode is the electrode of interest) or the internal resistance of the set-up.
In this chapter, a potentiostatic mini-scale bioelectrochemical reactor was designed to
study in details the biofilm-anode interaction. The method used here tries to evaluate redox
activity not only in the boundary layer of the electrode but also mediators that are located deeper
inside the biofilm. This is accomplished by using ―ultra‖ low scan rates (0.01 mV s-1
) which
enables a more pronounce influence of the electrode potential on redox active species (mediators)
that are not readily accessible by the electrode (i.e. far away from the electrode, confined within
the biofilm or inside the biomass).
Chapter 5: An Anodophilic Biofilm prefers a Low Anodic Potential
- 85 -
5.2 Experimental Section
5.2.1 Anodophilic Biofilm and Growth Medium
The electrochemically active anodophilic biofilm used in this study was established from
an acetate-enriched MFC that had been operated for over 200 days. Detail operation of the MFC
was described in Chapter 3. In brief, the biofilm was derived from an activated sludge inoculum
obtained from a local domestic wastewater treatment plant. During MFC operation, the anodic
potential lied mostly in a range between -300 and -450 mV vs. Ag/AgCl. At day 216, about two-
thirds of the biofilm coated anode granules were harvested from the MFC and were immediately
immersed in a growth medium to preserve the bacterial viability before they were packed into
the bio-electrochemical cell used in this study. Excess biofilm covered granules were stored in
the medium at 4 oC for later use. The growth medium consisted of (mg L
-1): NH4Cl 125,
NaHCO3 125, MgSO4·7H2O 51, CaCl2·2H2O 300, FeSO4·7H2O 6.25, and 1.25 mL L-1
of trace
element solution, which contained (g L-1
): ethylene-diamine tetra-acetic acid (EDTA) 15,
ZnSO4·7H2O 0.43, CoCl2·6H2O 0.24, MnCl2·4H2O 0.99, CuSO4·5H2O 0.25, NaMoO4·2H2O
0.22, NiCl2·6H2O 0.19, NaSeO4·10H2O 0.21, H3BO4 0.014, and NaWO4·2H2O 0.050. The
medium was complemented with 50 mM phosphate buffer to maintain a constant pH of 7.
5.2.2 Construction and Operation of Bio-electrochemical Cell
Figure 5.1 illustrates the schematic diagram of the three-electrode bio-electrochemical
cell used in this study. Similar to a conventional two-chamber MFC, it consisted of a working
and a counter chamber. The biofilm covered electrode was placed inside the working chamber,
which could accommodate a volume of about five cubic centimeters of the graphite granules.
Electrical contact between the graphite granules and the potentiostat was secured by using a
graphite rod current collector (5 mm diameter), which was embodied and pierced through a
rubber cover lid of the working chamber. A platinum sheet and a silver-silver chloride reference
electrode (3 M KCl) served as the counter and reference electrodes, respectively.
To minimize mass transfer limitation inside the working chamber, the working medium
was recirculated over the granular graphite at a flow rate of about 10 mL·min-1
by using a
peristaltic pump (Cole-Parmer Masterflex, model 7519-00). The recirculation tubing was
impermeable to oxygen (Tygon® , lumen diameter 3 mm) and was wrapped with aluminium foil
to avoid growth of photo-heterotrophic bacteria. The counter electrolyte was 100 mM phosphate
Chapter 5: An Anodophilic Biofilm prefers a Low Anodic Potential
- 86 -
buffer solution (pH 7). It was continuously mixed by using a magnetic stirrer bar and was
regularly renewed throughout the potentiostatic study.
Figure 5.1 (A) Schematic diagram and (B) photo of the mini-scale bio-electrochemical system used for
the potentiostatic experiments. (1) Graphite rod as current collector (Five millimeters diameter);
(2) silver-silver chloride reference electrode; (3) Biofilm coated graphite granules; (4) Separator
membrane (Nafion 117 proton exchange membrane); (5) Platinum foil; (6) 100 mM phosphate
buffer (pH 7); (7) Magnetic stirrer bar.
Since ultra slow potential scans of the working electrode were used in the study, it is
essential to avoid electrochemical contact between any electrochemically active species of the
V
i
LabVIEW™
Potentiostat
Counter Working Reference
InflowOutflow
(1 )(2 )
(4 )
(5 )
(6 )
(7 )
(3 )
V
i
LabVIEW™
Potentiostat
Counter Working Reference
InflowOutflow
(1 )(2 )
(4 )
(5 )
(6 )
(7 )
(3 )
(A)
(B)
Chapter 5: An Anodophilic Biofilm prefers a Low Anodic Potential
- 87 -
biofilm system and the counter electrode. Hence, instead of placing the three electrodes into a
common liquid medium, a proton exchange membrane (Nafion 117, Fuel cell store™) was used
to separate the biofilm covered working electrode as well as its associated growth medium from
the reference and the counter electrodes. This configuration was designed to simulate a MFC
operation with the use of a potentiostat to maintain a precise biofilm-electrode potential.
All experiments were conducted under well-controlled electrochemical conditions. Prior
to each experiment, the platinum counter electrode was pre-treated by heating in a flame to a
red-heat condition followed by quenching in deionized water. The biofilm electrode (i.e.
working electrode) was polarized against the reference electrode at desired potential value by
using a computer-programmable potentiostat (manufactured and quality assured by Murdoch
University Electronic Workshop).
A computer program written with LabVIEW™ (version 7.1 National Instrument) was
used to control and manipulate signals of the potentiostat. An analog input/ output data
acquisition board (DAQ) (Labjack™, UE9, 12-bits of resolution) was used as an analog-digital
converter to interface between the computer and the potentiostat. Voltage and current from the
potentiostat were recorded at fixed time intervals and the data were logged into an Excel
spreadsheet with a personal computer.
Accuracies of all the voltage signals from the DAQ were verified by using a high
precision digital multimeter (resolution 10 µV). Quality of the electrochemical data generated by
the system was satisfactorily assured by running cyclic voltammograms of a standard
ferricyanide solution (various concentrations complemented with 50 mM phosphate buffer, pH 7)
(data not shown). To avoid abiotic evolution of hydrogen or oxygen at the electrode during the
potentiostatic studies, the potential of the working electrode was only controlled in a narrow
range (anywhere between 0 and -0.6 V vs Ag/AgCl). This potential range was selected in order
to simulate the real operating anodic condition of MFC.
Chapter 5: An Anodophilic Biofilm prefers a Low Anodic Potential
- 88 -
5.3 Results and Discussion
To investigate electrochemically active species in solution cyclic voltammetry is usually
the method of choice. Most commonly, this method is based on using clean and defined
electrodes to oxidize and reduce dissolved mediators present in the vicinity of the working
electrode. However, in the case of bioelectrochemical systems such as microbial fuel cells
electron shuttling species are present with the biofilm growing on the electrode surface. In order
to evaluate potential electrochemically active species in the anodic compartment of the microbial
fuel cell described, cycling potentials were applied directly by using the biofilm attached
electrode as the working electrode in a bio-electrochemical cell (Figure 5.1).
Figure 5.2 Cyclic voltammograms of biofilm covered graphite granules from the anode of an
established MFC under electron donor (acetate) saturated and starved conditions. Scan
rate was 0.1 mV·s-1
.
5.3.1 Only Actively Metabolizing Biofilm Exhibited Electrochemical Activity in Cyclic
Voltammetry
When just using the biofilm covered graphite granules as the working electrode without
the supply of acetate, and employing CV at a scan rate of 0.1 mV·s-1
no significant
electrochemically active species could be detected (Figure 5.2). However, when the biofilm was
supplied with saturating concentrations of its electron donor (10 mM acetate) there was
background anodic current observed at most potentials in both scanning directions, suggesting
that the working electrode would serve as an anode for the biofilm (Figure 5.2). This
-0.10
0.00
0.10
0.20
0.30
-0.45 -0.35 -0.25 -0.15
Potential (V vs Ag/AgCl)
Cu
rre
nt (m
A) Acetate Saturated
Acetate Starved
Chapter 5: An Anodophilic Biofilm prefers a Low Anodic Potential
- 89 -
phenomenon is expected and has also been described by Fricke et al. (2008). Such anodic
current is not caused by the voltage applied by the potentiostat but by the bacteria, provided that
the electrode potential was held sufficiently high (> -380 mV vs. Ag/AgCl) to enable bacterial
transfer of electrons from acetate to the electrode.
Figure 5.3 Effect of acetate depletion on the current profile during repetitive cyclic voltametry
of the active, MFC enriched anodophilic biofilm covered graphite granules with a scan
rate of 0.1 mV·s-1
. Asterisks indicate the characteristic current peaks observed.
The peaks in the cyclic voltammogram obtained under substrate saturation (Figure 5.2)
suggest the presence of an electrochemically active species in the biofilm with potentials ranging
from -320 to -220 mV vs Ag/AgCl. To verify this effect, a limiting amount of the electron donor
of the biofilm (acetate) was added to the cell and repetitive cycles of CV at 0.1 mV·s-1
were run.
Results showed that initially the peaks of the CV were as described above and could be shown
as current peaks over time (Figure 5.3). As the acetate concentration diminished, the current
measured during the cycles changed by:
the decrease in overall current as the bacteria no longer had an electron donor.
the disappearance of the redox peaks obtained (depicted as asterisks in Figure 5.3).
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0 2 4 6 8 10 12 14
Time (hour)
Cu
rre
nt (m
A)
-0.45
-0.4
-0.35
-0.3
-0.25
-0.2
-0.15
Po
ten
tial (V
vs. A
g/A
gC
l)
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0 2 4 6 8 10 12 14
Time (hour)
Cu
rre
nt (m
A)
-0.45
-0.4
-0.35
-0.3
-0.25
-0.2
-0.15
Po
ten
tial (V
vs. A
g/A
gC
l)
Acetate sufficient Acetate diminishing
Cycle: 1st 2
nd 3
rd 4
th 5
th 6
th 7
th
Chapter 5: An Anodophilic Biofilm prefers a Low Anodic Potential
- 90 -
After addition of acetate to the depleted cell, the previous characteristic peaks and current
levels resumed. To test whether the peaks in the cyclic voltammograms were caused by potential
mediators present in the bulk liquid, the bulk liquid was drained and replaced by fresh medium.
This had no effect on the shape of the cyclic voltammogram (data not shown), supporting the
findings obtained by Fricke et al. (2008) in a Geobacter bioelectrochemical cell showing that the
redox active species are predominately located inside the biofilm or as part of the bacterial cells.
5.3.2 Redox Mediators of the Active MFC Biofilm were Located Far Away from the
Electrode Surface
When applying CV on the acetate saturated and hence actively current generating cell, a
reproducible current-potential pattern occurred (Figure 5.4). However, at scan rates commonly
used for detecting mediators in MFC systems (i.e. ranging from 5 to 50 mV·s-1
) (He et al. 2005;
Logan et al. 2006; Qiao et al. 2008; Rabaey et al. 2004), no particular peaks of electrochemically
active species could be detected (Figure 5.4). This could be due to the fact that the
electrochemically active species in the biofilm may not be in the immediate vicinity of the
electrode (i.e. the boundary layer) as it is known for CV with clean and defined electrodes.
Figure 5.4 Cyclic voltammograms of an active, acetate saturated biofilm covered graphite
granules (anode) of an established MFC with (A) scan rates commonly used for
conventional electrochemical study of mediator in MFCs and (B) very low scan rates.
It was considered that the scan rate should be further reduced to allow the applied
potentials to affect mediators further away from the electrode potentially situated within the
-3
-2
-1
0
1
2
3
-0.55 -0.45 -0.35 -0.25 -0.15 -0.05 0.05
Potential (V vs Ag/AgCl)
Curr
ent
(mA
)
5 mV/s10 mV/s25 mV/s50 mV/s
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
-0.45 -0.4 -0.35 -0.3 -0.25 -0.2 -0.15
Potential (V vs. Ag/AgCl)
Curr
ent
(mA
)
0.2 mV/s0.4 mV/s0.8 mV/s
(A) (B)
Chapter 5: An Anodophilic Biofilm prefers a Low Anodic Potential
- 91 -
biofilm or even inside the bacterial cells. Therefore, the scan rate was further reduced to an
extremely low rate of 0.01 mV·s-1
with the same biofilm. Although, this ultra low scan rate was
rarely reported in the literature, this could offer a longer equilibration time for the biofilm to
response to the effect of electrode potential.
Clearly, such an extremely slow scan revealed the presence of two (instead of one)
reproducible and distinct peaks at about -310 and -240 mV vs Ag/AgCl, respectively (Figure 5).
The two peaks obtained during the increasing potential scan (i.e. upper curve) signified either (a)
the discharge of mediators that the bacteria had previously reduced at lower potentials; or (b)
increased capacity of bacteria to produce electricity at higher rates at these higher potentials.
Figure 5.5 Ultra slow cyclic voltammetric scan (0.01 mV·s-1
) of an actively acetate
metabolizing MFC biofilm.
5.3.3 Detail Interpretation of the Anodophilic Properties of the MFC Biofilm in CV
In order to determine which of the above statement is more likely to explain the peaks
obtained, a more detailed scan was recorded around the voltage of the observed peaks. Upon
more close investigation of the electrochemical behavior of the biofilm by running scans around
the two key peaks at -310 mV and -240 mV vs Ag/AgCl, a key difference could be detected
between the two peaks (Figure 5.6).
-0.05
0.00
0.05
0.10
0.15
0.20
0.25
-0.55 -0.45 -0.35 -0.25 -0.15 -0.05 0.05
Potential (mV vs. Ag/AgCl)
Curr
ent
(mA
)
Peak 2 Peak 1
Chapter 5: An Anodophilic Biofilm prefers a Low Anodic Potential
- 92 -
Figure 5.6 Cyclic voltammetry with a narrow potential range covered (A) around the first peak
at -310 mV vs. Ag/AgCl and (B) around the second peak at -230 mV vs. Ag/AgCl. Scan
rate was 0.01 mV·s-1
.
The peak at -310 mV vs. Ag/AgCl was only observed during upward scans while
downward scans showed no such peak (Figure 5.6A). Instead the downwards scan showed a
gradual decrease in current caused by the deceleration of bacterial electron transfer rate. This
peak is more in line with the traditional view that a maximum in electron flow is caused by the
discharging of previously charged electron mediators as also observed by other authors (Fricke
et al. 2008).
0.00
0.05
0.10
0.15
0.20
0.25
0.30
-0.33 -0.32 -0.31 -0.30 -0.29 -0.28 -0.27
Potential (V vs. Ag/AgCl)
Cu
rre
nt (m
A)
0.10
0.11
0.12
0.13
0.14
0.15
0.16
0.17
0.18
-0.26 -0.25 -0.24 -0.23 -0.22 -0.21
Potential (V vs. Ag/AgCl)
Cu
rre
nt (m
A)
(A) 1st Peak:
(B) 2nd
Peak:
Chapter 5: An Anodophilic Biofilm prefers a Low Anodic Potential
- 93 -
In contrast, the peak at -230 mV vs. Ag/AgCl was visible in both scanning directions
(Figure 5.6B). This phenomenon is not in line with typical CV results. It cannot be explained by
the charging and discharging of electrochemically active compounds by the electrode. This
suggests that this second peak is not due to the redox behavior of mediator but that it is caused
by a maximum in bacterial activity at the potential of about -230 mV, independent on whether
previously higher or lower electrode potentials were applied.
5.3.4 Step-Change of Electrode Potential Signified the Existence of an Optimal
Electrode Potential for Current Production from the Biofilm
If the peak at -230 mV vs. Ag/AgCl observed above is truly due to the microbes showing
increased metabolic activity at this potential, then this maximum should also appear under
steady state conditions at a fixed potential. In order to verify whether there is such an optimum
electrode potential at which the anodic current shows a maximum, the MFC was run at fixed
potentials for extended periods of 1 h, before step-changing to a different potential, which was
again kept for 1 h (Figure 5.7).
Figure 5.7 Microbially generated current flow with acetate as electron donor at different
electrode potentials in a constant potential step-change experiment (1 h per step).
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
0.18
0 2 4 6 8 10 12 14
-0.34
-0.32
-0.30
-0.28
-0.26
-0.24
-0.22
-0.20
-0.18
Current
Potential
Time (h)
Curr
ent (m
A)
Pote
ntia
l (V v
s A
g/A
gC
l)
Peak 4 Peak 1 Peak 3
Peak 2
Chapter 5: An Anodophilic Biofilm prefers a Low Anodic Potential
- 94 -
The resulting current profile confirmed that the biofilm always gave a higher current
between -220 and -240 mV vs. Ag/AgCl than at either –180 to -200 mV vs. Ag/AgCl or at < -
260 mV vs. Ag/AgCl. Such current maximum remained when step changing the potentials either
upwards or downwards (Figure 5.8).
Figure 5.8 Steady state current as a function of electrode potential (data from Figure 5.7). Note:
the average current was calculated from data obtained in the last 30 min of the respective
potential.
Figure 5.9 Microbially generated current flow with acetate as electron donor at three applied
electrode potentials across the second peak at -230 mV vs. Ag/AgCl under steady state
conditions (1 h constant potential).
0.00
0.05
0.10
0.15
0.20
-0.35 -0.30 -0.25 -0.20
Potential (mV vs. Ag/AgCl)
Cu
rre
nt (m
A)
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0 1 2 3 4 5 6 7
Time (hour)
Cu
rre
nt
(mA
)
-0.255
-0.245
-0.235
-0.225
-0.215
-0.205
-0.195
-0.185
Po
ten
tial (V
vs. A
g/A
gC
l)
Current
Potential
Chapter 5: An Anodophilic Biofilm prefers a Low Anodic Potential
- 95 -
In contrast, the peak at -310 mV vs. Ag/AgCl could only be seen during stepwise
increasing potentials (Figure 7, 0 – 3 h), while decreasing potentials (Figure 7, 11 – 14 h) did not
show such a peak. Again the results suggest that the two peaks observed are of a different nature.
The dependency of the current on the potential changes shown above was very
reproducible in the cell tested. In a final attempt to test whether the MFC biofilm truly produced
a higher current at a lower and hence for the bacteria less attractive electrode potential, again the
results showed that indeed at anodic potentials between - 225 and -250 mV vs. Ag/AgCl the
bacteria transferred electrons to the anode faster than at -190 mV vs. Ag/AgCl (Figure 5.9).
5.3.5 Implication of Findings
This chapter suggests that with cyclic voltammetry the electrochemical activity of a
MFC biofilm could only be detected when the biofilm was actively metabolizing and generating
an electron flow, but not when the same bacteria were starved of their electron donor. This is not
in line with the idea that the mediators indicated by the peak in the cyclic voltammogram
(Figure 5.2) were oxidized and reduced by the electrode directly. The results can be explained
by the concept that the mediators detected in this ―biochemical CV‖ were unable to be reduced
by the electrode and were strictly dependent on active bacterial metabolism. While it is clear that
this species actively participates in the electron transfer from bacteria to the electrode, it is not
likely that it is the terminal electron carrier as it would by definition be able to be re-oxidised by
the graphite electrode itself.
The possibility of an optimum anodic potential at which the bacteria can cause a
maximum of current could be significant during optimization of MFCs. In the literature there
were indications of such an optimum to exist (Aelterman et al. 2008; Cheng et al. 2008; Torres
et al. 2007), however it has not been identified or quantified as the current study does. In terms
of microbial activity and of thermodynamics it would be expected that the highest potential of
the electrode would enable the fastest electron flow from the bacteria to the electrode. The
reason for the presence of a particular electrode potential of about -230 mV vs. Ag/AgCl beyond
which the microbial current generation decreases is unclear. However, the measured electron
flow (current) only measures the rate at which acetate is being oxidized (dissimilation). Under
growing conditions a proportion of the acetate taken up is not dissimilated but assimilated which
does not result in a flow of electrons. Accordingly it is perceivable that if the acetate intake rate
Chapter 5: An Anodophilic Biofilm prefers a Low Anodic Potential
- 96 -
becomes limiting and at increasing potentials the generation of ATP from respiration becomes
more readily possible that a lowering of electron flow signifies an increase in the metabolic
quotient (i.e. respiration per unit biomass) of the cells.
On the other hand, it can be envisaged that the established biofilm (which was
predominately cultivated at a low operating anodic potential in the previous experiments, < -300
mV vs. Ag/AgCl) may have discontinued to synthesize internal electron carriers that are
responsible for shuttling electrons to a high potential electron acceptor (at which the biofilm had
rarely exposed to), but rather the bacteria had switched their metabolism in synthesizing the
electron carrier(s) that are more thermodynamically compatible with (or had a higher affinity to)
a specific, more negative anodic potential. In nature, this situation would be similar to
facultative bacteria that can stop expressing electron carriers required for oxygen reduction
under anaerobic condition (Van Spanning et al. 1995). For example, a facultative nitrate-
reducing bacterium Paracoccus denitrificans was found to denitrify (nitrate) even at an oxygen
level of 100% air saturation (Sears et al. 1997). Here, the question of whether the anodophilic
biofilm can adapt their electron shuttling machinery in favor of the redox potential of their
associated electrode still remains open. Yet, this question could be valuable not only to BES
application and also to fundamental microbiology.
Further researches are warranted to elucidate in detail the electron transfer mechanism
(preferably at a protein-level) of the observed dependency between the electrode-living biofilm
and its surrounding redox environment.
- 97 -
6 Alternating Bio-Catalysis of Anodic and Cathodic
Reactions alleviates pH Limitation in a BES
— An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction —
(This chapter has been published in Environ. Sci. Technol. (2010) Vol.44(1)518-525)
Chapter Summary
As stated in Chapter 1, the poor cathodic oxygen reduction and the detrimental build-up
of electrolyte pH are fundamental hurdles in sustainable microbial fuel cell (MFC) development.
This chapter describes and tests a concept that can help overcoming both of these limitations, by
inverting the polarity of the MFC repeatedly, allowing anodic and cathodic reaction to occur
alternately in the same half-cell and hence neutralizing its respective pH effects.
For simplicity, we studied polarity inversion exclusively in one half cell, maintaining its
potential at -300 mV (vs. Ag/AgCl) by a potentiostat. The alternating supply of acetate and
dissolved oxygen to the biofilm resulted in the tested half cell to turn repeatedly from anode to
cathode and vice versa. This repeated inversion of current direction avoided the detrimental
drifting of the electrolyte pH. Control runs without current inversion ceased to produce current,
due to anode acidification.
The presence of the anodophilic biofilm survived the intermittent oxygen exposure and
could measurably facilitate the cathodic reaction by reducing the apparent oxygen overpotential.
It enabled the cathodic oxygen reduction at about -150 mV (vs. Ag/AgCl) compared to -300 mV
(vs. Ag/AgCl) for the same electrode material (granular graphite) without biofilm. Provided a
suitable cathodic potential was chosen, the presence of ―anodophilic bacteria‖ at the cathode
could enable a 5-fold increase of power output.
Overall, the ability of an electrochemically active biofilm to catalyze both substrate
oxidation and cathodic oxygen reduction in a single bioelectrochemical system has been
documented. This property could be useful to alleviate both the cathodic oxygen reduction and
the detrimental drifting of electrolyte pH in a MFC system. Further research is warranted to
explore the application of such bidirectional microbial catalytic property for sustainable MFC
processes.
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 98 -
6.1 Introduction
Microbial fuel cells (MFCs) represent a bio-electrochemical process that can directly
convert chemical energy stored in organic substances into electrical energy. In theory, any
biodegradable organic substance can potentially serve as the ―fuel‖ for this process. Therefore,
MFCs are seen as a promising waste-to-energy environmental technology. Yet, there are many
bottlenecks that hinder their development as a sustainable energy conversion process. Apart
from other upscale technical hurdles such as material costs and reactor configurations, the poor
cathodic oxygen reduction and the establishment of a pH gradient between anodic and cathodic
regions are currently perceived as the two major fundamental hurdles in MFC development
(Logan 2008; Rabaey et al. 2008; Rozendal et al. 2006a; Zhao et al. 2006a).
Continuous operation of MFC causes acidification at the anode due to the accumulation
of protons liberated from the microbial oxidation of organic compounds (eq. 6-1), whereas
alkalinization at the cathode is observed due to continuous consumption of protons by the
oxygen reduction reaction (eq. 6-2) (Zhao et al. 2006a). Given that oxygen accepts four
electrons, then four moles of protons are required for each mole of oxygen molecule used.
CH3COO- + 4 H2O 2 HCO3
- + 9 H
+ + 8 e
- (eq. 6-1)
2 O2 + 8 H+ + 8 e
- 4 H2O (eq. 6-2)
Ideally, these protons are replenished from the anodic oxidation of organic substrate. However,
ionic species such as sodium, potassium, ammonium, magnesium, and calcium will
preferentially migrate across a cation membrane as their concentrations are commonly much
higher (Rozendal et al., 2006). Consequently, the continuous consumption of protons for the
oxygen reduction reaction results in a net increase of catholyte pH, while the anode will become
acidified due to accumulation of protons. These lead to a so-called membrane pH gradient
overpotential or pH splitting limitation, which puts an electrochemical/ thermodynamic
constrain on the MFC performance (Harnisch et al. 2008; Rabaey et al. 2008; Rozendal et al.
2008a).
Harnisch et al. (2008) have recently highlighted the proton migration mechanism and
problems associated with the use of proton exchange-, cation exchange-, anion exchange- and
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 99 -
bipolar membranes in MFC systems. They concluded that none of these membrane separators
could circumvent the pH splitting problem, which leads to diminished bacterial activity and poor
cathodic oxygen reduction. To rectify the problem, active pH control such as acid/ base dosing
or addition of chemical buffer (e.g. phosphate buffer or bicarbonate) are required to sustain
steady current production. However, these approaches are not practical as they represent a
significant cost and energy input (Harnisch and Schroder 2009; Pham et al. 2009; Rozendal et al.
2008a).
Another key hurdle of MFC technology is the poor cathodic reaction when oxygen is
used as the terminal electron acceptor. Although oxygen is considered to be the most sustainable
terminal electron acceptor for MFC, the poor kinetics of oxygen reduction (i.e. high cathodic
overpotential) under conditions prevailing in most MFC configurations (i.e. neutral electrolyte
pH, low electrolyte ionic strength and ambient temperature) diminishes the performance of
oxygen cathode (Freguia et al. 2007b; Yu et al. 2007). The use of platinum as catalyst is not seen
as a sustainable solution because of costs and environmental impact in its production (Freguia et
al. 2007b; He and Angenent 2006). Recent studies show that bacteria can play a role in the
cathodic oxygen reduction (Clauwaert et al. 2007a; Freguia et al. 2007c; Rabaey et al. 2008;
Rozendal et al. 2008b) and promise to become an alternative to the conventional approach.
However, their potential is not fully explored.
Recently, it has been shown that the catalytic activity of an anodophilic biofilm is not
compromised by the presence of oxygen (Biffinger et al. 2008; Malki et al. 2008). Whether an
anodophilic biofilm could also assist oxygen reduction at a cathode has not been investigated so
far. If a biofilm could catalyze both anodic substrate oxidation and cathodic oxygen reduction,
an alternating current would be obtained when exchanging substrate and oxygen supply from
one half cell to the other. Since the anodic substrate oxidation is a proton liberating process
while the cathodic oxygen reduction is a proton consuming process, such inversion of MFC
polarity may potentially alleviate the pH limitation during MFC operation.
The current study aims at addressing the problem of pH gradient as well as the slow
cathodic oxygen reduction by intermittently reversing the polarity of the MFC. The specific
research questions are:
1) Can the detrimental build-up of electrolyte pH be avoided?
2) Can anodophilic bacteria catalyze the cathodic oxygen reduction?
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 100 -
As a proof-of-concept study, a potentiostatically controlled two-chamber MFC was used
to answer the above two questions.
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 101 -
6.2 Experimental Section
6.2.1 Electrochemically Active Biofilm and Growth Medium
The electrochemically active biofilm used in this study was enriched in the anodic
chamber of an acetate-fed MFC during continuous operation for over 200 days (Chapter 3,
Cheng et al. 2008). The initial inoculum before that time was activated sludge. During MFC
operation, the anodic potential was between -200 and -450 mV vs. Ag/AgCl. The growth
medium consisted of (mg L-1
): NH4Cl 125, NaHCO3 125, MgSO4·7H2O 51, CaCl2·2H2O 300,
FeSO4·7H2O 6.25, and 1.25 mL L-1
of trace element solution, which contained (g L-1
): ethylene-
diamine tetra-acetic acid (EDTA) 15, ZnSO4·7H2O 0.43, CoCl2·6H2O 0.24, MnCl2·4H2O 0.99,
CuSO4·5H2O 0.25, NaMoO4·2H2O 0.22, NiCl2·6H2O 0.19, NaSeO4·10H2O 0.21, H3BO4 0.014,
and NaWO4·2H2O 0.050. Yeast extract solution was periodically added (every 4 days) to the
medium (50 mg L-1
final concentration) as bacterial growth supplement. This medium was used
as the electrolyte in the working chamber throughout the entire study. A similar medium but
complemented with 50 mM potassium phosphate buffer (pH 7) was used as the counter
electrolyte. Both the working and counter electrolytes were renewed regularly.
6.2.2 Construction and Monitoring of the Bioelectrochemical System
Figure 6.1 illustrates a schematic diagram of the electrochemical cell and its peripheral
settings. A two chamber electrochemical cell with identical electrode material was used as
described in Chapter 2-4. The two chambers were of equal volume and dimension (316 mL (14
cm 12 cm 1.88 cm)). They were physically separated by a cation-selective membrane (CMI-
7000, Membrane International Inc.) with a surface area of 168 cm2. Both chambers were filled
with identical conductive granular graphite (El Carb 100, Graphite Sales, Inc., USA, granules 2-
6 mm diameter). This reduced the void volume of the working chamber from 316 to 160 mL. A
total volume of 350 mL of the electrolyte was re-circulated through the working half cell.
The electrode covered with the anodophilic biofilm described above was defined as the
working electrode, while the other served as the counter electrode. The working electrode was
polarized against a silver-silver chloride reference electrode (saturated KCl) at defined potential
values by using a computer-controllable potentiostat (manufactured and quality assured by
Murdoch University Electronic Workshop). The reference electrode was placed inside the
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 102 -
working chamber, with a distance between the reference electrode and the anode of less than one
cm. This is in contrast to the previous chapters (Chapter 4 & 5, Cheng et al. 2008) where the
reference electrode was 15 cm away from the working electrode, resulting in an average of about
40 mV lower potential readings. This difference has been taken into account in Figure 6.7A. All
electrode potentials (mV) described in this article refer to values against Ag/AgCl reference
electrode (3 M KCl, RE-5B, Bioanalytical Systems, Inc., West Lafayette, Indiana, USA; ca.
+197 mV vs. standard hydrogen electrode (Bard and Faulkner 2001)).
Figure 6.1 Schematic diagram of the potentiostat-controlled two-chamber bioelectrochemical
system used in this study. RE = reference electrode; WE = working electrode; CE =
counter electrode.
A computer program LabVIEW™ (version 7.1 National Instrument) was developed to
continuously control and monitor the bioprocess. An analog input/ output data acquisition card
(Labjack™, U12) was used as an analog-digital converter to interface between the computer and
the potentiostat. Voltage signals from all the monitoring devices were recorded at fixed time
intervals via the LabVIEW software interfaced with a National InstrumentTM
data acquisition
card (DAQ). All data were logged into an Excel spreadsheet. Accuracies (less than 0.1 mV) of
Moisten Air
Stirrer
DAQ Board
Computer
Control/
Monitoring by
LabVIEW™
Substrate Feed
Moisten
Air
pH Eh pH
Ag/AgCl
Reference
Electrode
Biofilm-Working
Electrode
(Granular Graphite)
Counter Electrode
(Granular Graphite)
Cation Exchange Membrane
Recirculation
Pump
Counter
Electrolyte
Graphite Rod
( Diameter = 5mm)
Out In
Aquarium
Air Pump
DI Water
PC-ControllablePotentiostat
RE WE CE
One Way
Exhaust
DO
DO
Moisten Air
StirrerStirrer
DAQ Board
Computer
Control/
Monitoring by
LabVIEW™
Substrate Feed
Moisten
Air
pH Eh pH
Ag/AgCl
Reference
Electrode
Biofilm-Working
Electrode
(Granular Graphite)
Counter Electrode
(Granular Graphite)
Cation Exchange Membrane
Recirculation
Pump
Counter
Electrolyte
Graphite Rod
( Diameter = 5mm)
Out In
Aquarium
Air Pump
DI Water
PC-ControllablePotentiostat
RE WE CE
One Way
Exhaust
DO
DO
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 103 -
all the voltage signals from the DAQ were verified periodically by using a high precision digital
multimeter (resolution 10 µV).
Oxidation-reduction potential (Eh), pH and dissolved oxygen (DO) concentration
(Mettler Toledo, InPro6800; 4100 PA D/O Transmitter) were continuously monitored at various
locations as specified in Figure 1. Acetate concentration was determined by using a gas
chromatography (GC) as described (Cheng et al. 2008).
6.2.3 Operation of the Bioelectrochemical Process
6.2.3.1 General Operation
To minimize mass transfer limitation, both electrolytes were continuously agitated by
pumping (8 L h-1
) through a separate recirculation loop. The supply of acetate and oxygen
(moist atmospheric air) to the working chamber was computer controlled. A peristaltic pump
(Cole-Parmer, Masterflex®
C/L®
) and an aquarium air pump were used to deliver the acetate
feed (1 M) and air via the inflow of the working chamber at a defined flow rate, respectively. All
experiments were conducted at 30 ± 2 oC.
6.2.3.2 Alternating Supply of Acetate and Oxygen to the Anodophilic Biofilm
The proposed process consists of two phases: (I) an anoxic substrate supply phase and (II)
an aerobic oxygen supply phase (Figure 6.2). During both phases the same electrolyte was used.
In Phase I acetate was added as a spike of 0.5 mmole (resulting in 1.4 mM to simulate
the concentration of a wastewater) to the working chamber, in which the electrochemically
active biofilm attached to the working electrode oxidized the acetate using the working electrode
as the electron acceptor. Acetate was added as a spike rather than continuously fed to allow its
depletion prior to the next phase where the electrolyte was oxygenated, as the presence of both
acetate and oxygen will enable their direct use by the bacteria resulting in less current output of
the cell (Oh et al. 2009).
After 2.5 hours oxygen was introduced into the working electrolyte commencing Phase
II. At this stage acetate had been shown to be consumed as determined by gas chromatography.
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 104 -
The lack of acetate towards the end of Phase I was also indicated by a characteristic drop in
current from typically above 200 mA to close to zero (Figure 6.4A). The oxygen supply to the
electrolyte caused the desired change in polarity, as the potentiostatically controlled working
electrode now became the cathode and resulting in negative current. The DO was online
monitored to ensure that it did not become limiting. Oxygen uptake rate (OUR) measurements
could be derived from online monitoring of the DO at the inflow and outflow as described below.
After 0.5 hours of operating with continuously aerated working electrolyte, the polarity
of the cell was reversed again to its original polarity by stopping oxygen supply and injecting 1.4
mM of acetate to the same working electrolyte.
Figure 6.2 Conceptual diagram of intermittent supply of electron donor and acceptor to an
anodophilic biofilm in a two-chamber potentiostatic microbial fuel cell.
Although the potential of the working electrode was maintained constant (-300 mV for
most experiments and +200 mV for Figure 6.5), the working electrode would serve as either an
anode or a cathode, depending on whether an anodic oxidation of electron donor (acetate) or a
cathodic reduction of acceptor (oxygen) was dominated at the working electrode. Overall, this
approach enabled the biofilm to neutralize the accumulated acidity or alkalinity in the electrolyte
on a sustainable basis. Hence, an active pH control of the electrolyte (e.g. acid/base dosing or
provision of extra pH buffer) was not needed.
Time
Action
Substrate Addition
Aeration Off
Action
Aeration On (O2 Supply)
Oxidation
CH3COO- +4H2O 2HCO3
- +
9H+ + 8e
-
Reduction
2O2 + 8H+ + 4e
- 4H2O
PHASE I (Anaerobic) PHASE II (Aerobic)
Effects
Anodic Current
pH Drop
Effects
Cathodic Current
pH Rise
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 105 -
6.2.3.3 Cathodic Electron Balances
The continuous monitoring of the DO concentrations (DO, mmol O2 L-1
) in the inflow
(DOin) and outflow (DOout) of the working chamber allowed On-line estimation of the OUR at
the working electrode. By considering the flow rate (FR, L h-1
) of the working electrolyte, the
OUR in Phase II could be deduced according to eq. 6-3.
FR
DODO inout )()hO (mmol OUR 1-
2
(eq. 6-3)
4OUR)h e (mmolreduction Ofor rate flowElectron -1-
2 (eq. 6-4)
The OUR is further converted into its corresponding electron flow rate according to eq. 6.4. By
integrating this electron flow rate and the actual cathodic electron flow rate (eq. 6-5) over time,
allowed establishing of a charge balance of the cathodic reaction (i.e. cathodic coulombic
efficiency). Where I is the current (mA) and F is Faraday‘s constant (96,485 C mol e-1
).
F
I3600)h e (mmolcurrent cathodic as rate flowElectron 1-- (eq. 6-5)
6.2.4 Catalytic Effect of the Anodophilic Biofilm on Cathodic Oxygen Reduction
A separate experiment was conducted to establish the relationship between the cathodic
current and electrode potentials. Prior to the experiment, the working chamber was flushed with
fresh electrolyte in order to remove any residual acetate in the system. Thereafter, the electrolyte
was continuously aerated to obtain DO of >3 mg L-1
. Steady state cathodic currents were
recorded and plotted against the electrode potentials ranging from -50 to -400 mV. Abiotic
control experiments were performed using abiotic plain graphite granules as the working
electrode, while all other operating parameters were identical to those in the biotic settings. At a
given potential, the differences in reaction velocity (here current) caused by the biofilm were
considered as catalysis.
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 106 -
6.3 Results and Discussion
6.3.1 Acidification of Electrolyte diminishes Anodic Current Production
At an electrode potential of -300 mV, the addition of acetate could immediately initiate
an anodic electron flow, yielding an anodic current peak of about 240 mA (1500 A·m-3
void
working chamber volume) (Figure 6.4A). Upon acetate depletion (data not shown), the current
started to decline to the original background level.
Figure 6.3 Effect of acetate spikes on current in a potentiostatic microbial fuel cell: (A) with
electrolyte pH controlled at 7; and (B) without pH control of electrolyte. Arrows indicate
acetate additions (450 moles per spike); biofilm-electrode was polarized at -300 mV;
dissolved oxygen in electrolyte was zero.
0
20
40
60
80
100
120
140
Cu
rre
nt (m
A)
4
5
6
7
pH
0
20
40
60
80
100
0 2 4 6 8 10 12 14Time (h)
Cu
rre
nt (m
A)
4
5
6
7
pH
CurrentpH
A.
B.
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 107 -
Figure 6.4 Effect of intermittent supply of acetate and oxygen on (A) current; (B) pH; (C)
electrolyte Eh; (D) inflow/outflow dissolved oxygen concentrations and (E) cathodic
electron flow rate calculated from OUR and current in the biofilm-contained working
chamber of a potentiostatic microbial fuel cell. Note: black arrows indicate acetate
addition; working electrode potential was poised at -300 mV throughout the experiment.
Since the microbial oxidation of acetate is a proton liberating process (see eq. 6-1), the
electrolyte pH decreased over time. Unless pH control is used, this decrease in pH eventually
6.5
6.6
6.7
6.8
6.9
7.0
7.1
pH
-150
-50
50
150
250
Cu
rre
nt (
mA
)
0
0.1
0.2
0.3
DO
(m
mo
l O2/
L)
.
Inflow
Outflow
0
2
4
6
8
0 2 4 6 8 10 12 14 16
Ca
tho
dic
Ele
ctr
on
Flo
w
Ca
lcu
late
d
(mm
ole
e- / h
) .
Time (h)
From OUR
From Cathodic Current
-300
-200
-100
0
100
200
300
Eh
(m
V v
s. A
g/A
gC
l)
6.5
6.6
6.7
6.8
6.9
7.0
7.1
pH
-150
-50
50
150
250
Cu
rre
nt (
mA
)
0
0.1
0.2
0.3
DO
(m
mo
l O2/
L)
.
Inflow
Outflow
0
2
4
6
8
0 2 4 6 8 10 12 14 16
Ca
tho
dic
Ele
ctr
on
Flo
w
Ca
lcu
late
d
(mm
ole
e- / h
) .
Time (h)
From OUR
From Cathodic Current
-300
-200
-100
0
100
200
300
Eh
(m
V v
s. A
g/A
gC
l)
A E R A T I O N
A E R A T I O N
A E R A T I O N
A E R A T I O N
A E R A T I O N
Acetate
Addition
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 108 -
impedes the microbial oxidation of acetate. Figure 6.3A shows that when the pH was strictly
controlled at a set-point of 7, the anodophilic biofilm catalyzed acetate oxidation over prolonged
times. At a poised potential of -300 mV maximum anodic currents of 118 6.7 mA and
coulombic recoveries of 83.3 2.8 % were obtained (Figure 6.3A). However, without the
control of electrolyte pH the continuous build-up of acidity in the electrolyte hampered the
current producing activity of the anodophilic biofilm (Figure 6.3B). At pH 5.2, acetate
addition did not result in a clear current response. A similar pH dependent behavior of the MFC
biofilm has also been reported by Clauwaert et al. (2007b).
6.3.2 Alternating Supply of Acetate and Oxygen to a Biofilm limits Anodic Acidification
It is the intent of this chapter to test whether the detrimental pH drop in the anode can be
prevented by turning the anode intermittently into cathode (inversion of polarity), by merely
replacing the supply of acetate (electron donor) with oxygen (electron acceptor). Rather than
swapping over the oxygen supply from one electrode against the acetate supply of the other (it is
difficult to interpret results as both electrodes suddenly change their behavior), only the working
electrode was studied when supplied alternately with acetate and oxygen. This was done by
maintaining its potential constant at -300 mV a voltage that is realistic for both anodic and
cathodic electrodes in MFC. The other (biofilm free) electrode served merely as counter
electrode.
Figure 6.4 shows the changes of current, pH, Eh, DO and cathodic electron flow of the
potentiostatic MFC over an extended period during which the biofilm received alternating
supply of acetate or oxygen. This alternating supply of electron donor (acetate) and electron
acceptor (oxygen) inverted the electrode repeatedly from anode to cathode and back to anode. In
contrast to the operation without polarity inversion (Figure 6.3B), acetate gave repeated current
signals on a sustainable basis (Figure 6.4A).
Since acetate oxidation is a proton-liberating reaction and the electrolyte had only a low
pH buffering capacity (see composition of the medium in Section 2.1), the liquid medium (here
the working electrolyte) became significantly acidified during each anoxic (anodic) phase. For
each acetate addition (0.5 mmole), a drop of about 0.3 units of pH was observed (Figure 6.4B).
This pH drift was counteracted by increased DO level which also initiated an increase in
electrolyte Eh from around -200 to +200 mV (Figure 6.4C and D). As the potentiostat
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 109 -
maintained the biofilm working electrode at a constant potential of -300 mV the oxygen reaction
with the electrode caused now a reverse electron flow (cathodic current), which - as expected -
consumes protons at this working electrode (Figure 6.4A).
Figure 6.5 Effect of intermittent supply of acetate and oxygen on (A) current and pH; (B)
inflow/outflow dissolved oxygen concentrations and (C) cathodic electron flow rate
calculated from OUR and current in the biofilm-contained working chamber of a
potentiostatic microbial fuel cell. Note: black arrows indicate acetate addition; working
electrode potential was poised at +200 mV throughout the experiment.
By intermittently reversing polarity, the addition of acetate spikes resulted in current
peaks for longer and giving an average anodic coulombic recovery (83.5 4.6 %) for each
5.0
5.5
6.0
6.5
7.0
-10
40
90
140
190
240
pH
Cu
rre
nt (
mA
)
Current
pH
0
0.1
0.2
0.3
DO
(m
mo
l O2/
L)
Inflow
Outflow
0
2
4
6
0 5 10 15
Ca
tho
dic
Ele
ctr
on
Flo
w
Ca
lcu
late
d
(mm
ole
e- / h
)
Time (h)
From OUR
From Cathodic Current
5.0
5.5
6.0
6.5
7.0
-10
40
90
140
190
240
pH
Cu
rre
nt (
mA
)
Current
pH
0
0.1
0.2
0.3
DO
(m
mo
l O2/
L)
Inflow
Outflow
0
2
4
6
0 5 10 15
Ca
tho
dic
Ele
ctr
on
Flo
w
Ca
lcu
late
d
(mm
ole
e- / h
)
Time (h)
From OUR
From Cathodic Current
5.0
5.5
6.0
6.5
7.0
-10
40
90
140
190
240
pH
Cu
rre
nt (
mA
)
Current
pH
0
0.1
0.2
0.3
DO
(m
mo
l O2/
L)
Inflow
Outflow
0
2
4
6
0 5 10 15
Ca
tho
dic
Ele
ctr
on
Flo
w
Ca
lcu
late
d
(mm
ole
e- / h
)
Time (h)
From OUR
From Cathodic Current
A I
R
A I
R
A I
R
A I
R
A I
R
A I
R
A.
B.
C.
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 110 -
acetate addition that was similar to that with pH control at 7 (see Figure 6.3A). These results
suggested that the anodophilic activity of the biofilm were not affected by the intermittent
presence of DO in the system.
6.3.3 Operating the Described MFC at +200 mV did not enable a Cathodic Electron
Flow
A potential of -300 mV is typical for the anodes of MFC but it is lower than what
cathodes are usually aimed to be. To test whether the above process can also operate at a higher
electrode potential, the experiment was repeated at a different potential of +200 mV (Figure 6.5).
In contrast to the experiment at -300 mV, no significant cathodic current was observed which is
in line with the general observation that cathodic oxygen reduction requires potentials lower
than +200 mV (Clauwaert et al. 2007b; Dumas et al. 2008b; Freguia et al. 2007c; He and
Angenent 2006; Rabaey et al. 2008). The presence of anodophilic bacteria could not make a
difference here.
Without an operating cathodic reaction, the intermittent aeration could not neutralize the
acidity of the working electrolyte. As a consequence, the acidity that was previously caused by
the anodic acetate oxidation of the same biofilm could not be compensated, resulting in a
gradual decrease of electrolyte pH and diminishing anodic activity of the bacteria (Figure 6.5A).
6.3.4 Cathodic Electron Balance
In analogy to the anodic coulombic efficiency, which balances the measured electron
flow against the amount of electron donor (acetate) reacted, a cathodic coulombic efficiency can
be determined from the ratio of electrons flowing away from the cathode and the amount of
electron acceptor (here oxygen) consumed. This cathodic electron balance showed that there was
always more oxygen used by the biofilm than could be accounted for by the current driven by
the potentiostat. Even with no current some background oxygen uptake activity was observed
(Figure 6.4E). This oxygen uptake is most likely due to cell maintenance respiration or the
oxidation of cell internal energy storage produced from acetate during phase 1. This is supported
by the observation that at cathodic potentials of +200 mV no cathodic current was generated at
the working electrode but significant oxygen was consumed (Figure 6.5C).
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 111 -
6.3.5 Catalytic Effect of the Anodophilic Biofilm on the Cathodic Reaction
The conclusions from the above experiments are that the cathodic reaction of the
working electrode coated with biofilm worked satisfactorily at -300 mV but not at +200 mV. In
order to establish in which range of cathodic potentials the described biofilm shows the greatest
catalytic effect, the cathodic current of the biofilm coated MFC electrode was compared to that
of a fresh, abiotic graphite electrode (Figure 6.6).
Figure 6.6 Effect of the anodophilic biofilm on the cathodic current obtained at different
electrode potentials: (A) current vs. electrode potential; (B) current enhancement factor (i
Bio/ i Abio) vs. electrode potential.
While the abiotic electrode showed the typical effect of oxygen overpotential with
significant current only being generated below -300 mV, the anodophilic biofilm enabled current
0
2
4
6
8
10
12
-450-400-350-300-250-200-150-100-50
Electrode Potential (mV vs. Ag/AgCl)
iBio/ i
Abio
-90
-80
-70
-60
-50
-40
-30
-20
-10
0
Cu
rre
nt (m
A)
Biofilm-Electrode
Abiotic-Electrode
A.
B.
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 112 -
to be produced at -200 mV, signifying a lowering of the cathodic overpotential (Figure 6.6A).
This observation has been described before for cathodophilic biofilms (Clauwaert et al. 2007b;
Freguia et al. 2007c; Rabaey et al. 2008). At cathodic potentials between -250 and -300 mV the
presence of anodophilic bacteria enabled up to 10 times higher current than the abiotic control
(Figure 6.6B). At potentials more negative than -300 mV, the effect of the anodophilic biofilm
on the cathodic electron flow became less significant owing to the increased abiotic cathodic
current (Figure 6.6B). However, the less negative potentials (>-300 mV) are more relevant for
the practical operation of MFC.
Figure 6.7 Effect of the anodophilic biofilm on the hypothetical MFC power production: (A)
current-electrode potential curves; (B) power curves. Notes: power and voltage values
were calculated from P = V* I, where V is the difference between anodic and cathodic
0
10
20
30
40
50
60
70
80
-450 -400 -350 -300 -250 -200 -150 -100
Cu
rre
nt (
mA
)
Electrode Potential (mV vs. Ag/AgCl)
Biofilm cathode
Abiotic cathode
Biofilm anode (refer to Cheng et al., 2008)
A.
0
5
10
15
20
25
30
0 10 20 30 40 50
Po
we
r D
en
sity (W
/m3)
Current (mA)
Biofilm cathode
Abiotic cathode
B.
Rectangle Area= Power
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 113 -
potentials; The dotted arrows extending from the dotted circles indicate the theoretical
maximal currents obtainable with the abiotic or anodophilie-catalyzed cathode.
6.3.6 Expected Power Production of MFC with Anodophilic Bacteria at both the Anode
and Cathode
The results above described the cathodic current produced by an anodophilic biofilm
coated electrode as a function of the cathodic potential. Combining this cathodic polarization
curve with the anodic polarization curve of the same cell as published earlier (Cheng et al. 2008)
allows to predict the maximum MFC power production (Figure 6.7A, B).
In an electricity producing MFC, the anodic current must be equal to the cathodic current.
Hence, the intersection of the anodic and cathodic polarization curves in Figure 7A displays the
maximum possible current (assuming zero resistance) that can be produced by a MFC with both
electrodes covered by the biofilm. For any given anodic and cathodic potential, the current and
power produced can be predicted from the rectangle areas. (V × I) as illustrated in Figure 7A.
This information also allows us to obtain the traditional power curve (Figure 6.7B). The
differences in abiotic and biofilm catalyzed cathodic polarization curves allow to estimate that,
in our case, about five times more power can be produced due to cathodic catalysis of the
biofilm.
6.3.7 Implication of the Findings
6.3.7.1 Potential Benefits of the Proposed Concept to developing Sustainable MFC
Processes
In the literature, Freguia et al. (2007c) had proposed a new operating regime to address
the limitations of pH membrane gradient and poor oxygen reduction. This was accomplished by
allowing an acidified anolyte effluent to serve as a continuous feed for the cathode. As such, the
excess protons from the anode could react directly in the cathodic oxygen reduction reaction.
Further, the residual amount of organic carbon (e.g. acetate) in the anolyte could also support the
growth of heterotrophic bacteria at the cathode. These cathode associated bacteria could further
oxidize the remaining organics in the wastewater and were capable of catalyzing the cathodic
oxygen reduction, resulting in an overall power increase without the need of an active pH
control.
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 114 -
However, as mentioned by these authors the oxygen consumption by bacteria at the
cathode in the presence of soluble organic carbons could compete for oxygen with the cathode
and diminish electricity generation (Freguia et al. 2007c). Further, the inevitable anodic biofilm
acidification would still be an issue in such a system (Rozendal et al. 2008a; Torres et al. 2008b).
In contrast to the above, our proposed MFC operating regime may offer the following
additional benefits:
The separation of acetate and oxygen addition phases may avoid the undesirable growth
of aerobic heterotrophic bacteria in the system, which may potentially out-compete the
cathodophilic bacteria in the biofilm community.
As protons are continuously liberated within an anodophilic biofilm upon substrate
oxidation, the current production can be limited by the rate at which the protons are
transported out of the biofilm (Torres et al. 2008b). Our concept of alternating flow of
electrons not only can alleviate such localized pH effect inside the biofilm but use the
protons accumulated from the anodic reaction to stimulate the subsequent proton
requiring cathodic reaction. This is because the protons are not required to be transported
out of the biofilm, but indeed the protons can reside within the biofilm for the subsequent
oxygen reduction reaction.
6.3.7.2 Anodophiles Catalyzing the Cathodic Reaction
Our observation that the same biofilm may play a role in the catalysis of both the anodic
and cathodic reaction is not entirely new. For example, Geobacter sulfurreducens, a widely
described anodophile (Bond and Lovley 2003; Dumas et al. 2008a; Fricke et al. 2008; Reguera
et al. 2006) was found to catalyze the electron transfer from the cathode to a suitable electron
acceptor (fumarate) (Dumas et al. 2008b; Gregory et al. 2004). However, fumarate is a common
electron acceptor for anaerobic bacteria such as Geobacter, and not a sustainable electron
acceptor in MFC.
In a recent study, Rabaey et al. (2008) have found that a mixed culture biofilm enriched
from an air-biocathode comprised mainly of Proteobacteria and Bacteroidetes species. A similar
microbial composition of an anodophilic biofilm was observed by Logan and Regan (2006a).
Such similarity suggests that perhaps the same groups of bacteria may catalyze both the anodic
and cathodic reactions.
Chapter 6: An Anodophilic Biofilm Catalyzes Cathodic Oxygen Reduction
- 115 -
Overall, the results in this chapter demonstrate for the first time that a mixed culture
anodophilic biofilm can also catalyze the cathodic reduction of oxygen in a single
bioelectrochemical system. The fact that an anodophilic biofilm that had developed over many
months of operation under strictly anaerobic conditions could sustain fully oxygenated
conditions without measureable side effects is significant and has also been observed recently by
Oh and coworkers (2009). The possible uses of a biofilm that catalyzes both the anodic and
cathodic reaction could lie in the operation of a traditional MFC by intermittent polarity
inversion (exchanging the electron donor and oxygen supply) at regular intervals. According to
the results of this chapter it should help overcoming the detrimental drifting of electrolyte pH
and poor cathodic oxygen reduction.
- 116 -
7 A Scalable Bioelectrochemical System for Energy
Recovery from Wastewater — Rotatable Bio-
Electrochemical Contactor (RBEC)
(A provisional patent has been filed for the invention described in this chapter)
Chapter Summary
Up to now, wastewater treatment using BES has not been successful on an industrial-
scale. This is partly due to the lack of a suitable reactor configuration for large-scale wastewater
application. In this chapter, a new BES configuration, termed rotatable bioelectrochemical
contactor (RBEC) is developed to convert the chemical energy from wastewater COD (acetate)
into electricity. Similar to rotating biological contactors (RBC), the RBEC consists of a
cylindrical water-holding vessel (ca. 3 L) which houses an array of discs (electrodes) mounted
onto a central horizontal rotatable shaft. Each disc consists of a water-immersed anodic and an
air-exposed cathodic half. No ion-exchange membrane and wastewater recirculation are required
as the air-water interface separated anode from cathode. Intermittent disc rotation (180°) enabled
each halve alternately serves as an anode and a cathode. This operating mode allows electricity
production without being affected by the pH limitation.
The COD removal caused by the action of the intermittently turning discs was increased
by about 15% (from 0.79 to 0.91 kg COD·m-3
·d-1
) by allowing an electron flow between the
anodic and cathodic disc halves. This result suggests that using suitable conductive material for
the disks of traditional RBC may already significantly improve the treatment performance.
Coupling the RBEC with a potentiostat to alleviate cathodic limitation and increase electron
flow could further increase COD removal rate to 1.32 kg COD m-3
day-1
(hydraulic retention
time 5h). About 40% of the COD degraded by the RBEC was via an electron flow meaning the
COD degrading bacteria were submersed and anoxic while the oxygen was consumed by the air
exposed cathodic disc half. While the COD removal rate was comparable to that of conventional
activated sludge processes (CAS), the RBEC removed COD more energy-efficiently than CAS
(0.47 vs. 0.7-2.0 kWh kgCOD-1
). Because of its novelty the system is expected to be improved
in both COD removal and energy use if more effective cathodic catalysts for MFC cathodes
become available.
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 117 -
7.1 Introduction
Chapter 6 has shown that a MFC anodophilic biofilm could alternately catalyze anodic
acetate oxidation and cathodic oxygen reduction reaction in a two chamber bioelectrochemical
cell with granular graphite used as the electrode material, causing a regular polarity inversion of
the system. This microbial property could alleviate both the pH splitting and poor cathodic
oxygen reduction limitations. However, it is not practical to implement the concept using the
two-chamber reactor configuration as it requires intermittent addition of a concentrated electron
donor substrate and the electron acceptor (i.e. oxygen via aeration) into the bulk liquid. This also
requires an extensive uplifting or re-circulating of the wastewater. Further, separation of anode
and cathode in the system require the use of sophisticated ionic exchange membranes, which in
general constitute a large portion of the capital cost of the whole system (Clauwaert et al. 2008a;
Rozendal et al. 2008a). Hence, eliminating the use of membranes can further reduce capital cost.
To simplify process up-scaling and to facilitate commercialization potential of BES, the
aforementioned limitations require not a small modification or process optimization but a
rethink of how the bacterial electrochemical activity can be better exploited with a more
practical and scalable BES reactor design. In this Chapter, a new BES configuration termed
rotatable bioelectrochemical contactor (RBEC) is developed. It can convert COD into electricity
without using conventional pH control methods (i.e. acid/base dosing or buffer addition) and
without the need of re-circulating the wastewater against gravity.
Similar to the well-known rotating biological contactor (RBC), the RBEC consists of a
cylindrical Perspex vessel (ca. 3 L), which houses 20 electrode discs that are mounted onto a
central horizontal rotating shaft. Each disc is divided into a water-submerged anodic and a
headspace-exposed cathodic half disc, which is partially immersed in the bulk liquid to allow
ionic conductance between the two electrodes. As such, the air-water interface served as a
―virtual membrane‖ separating the two redox zones eliminating the need of an ion-exchange
membrane. The electrodes of the RBEC are rotated such that a biofilm is established thereon and
alternately serves as the anode and the cathode by regularly swapping over the electrode discs.
The intermittent catalysis of anodic and cathodic reactions by a biofilm could help neutralize
acidity/ alkalinity of the biofilm.
The performance of the RBEC for electricity generation as a MFC is evaluated. The
feasibility of using an external power supply to facilitate the cathodic oxygen reduction and the
organic removal was also assessed. Energy consumption and treatment performances are
compared with the conventional AS process.
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 118 -
7.2 Experimental Section
7.2.1 Construction of the RBEC Reactor
The RBEC reactor consists of three major components: (1) a reactor vessel; (2) a
rotatable electrode disc assembly; and (3) a computer controllable stepper motor unit (Figure 7.1
Photo 7.1). The reactor vessel consists of a cylindrical Perspex transparent water pipe (298 mm
length; 140 mm diameter) with its two ends covered by two separate square Perspex side Photos
(160 x 160 mm). A rubber O-ring was placed between each side Photo and the centre pipe to
assure air and water tightness.
Figure 7. 1 A schematic diagram of the RBEC system: (A) operated as a microbial fuel cell for
electricity generation (batch operation); (B) coupled with a potentiostat for enhanced
cathodic oxygen reduction operated at a chemostat mode. (1) computer monitoring and
control using LabVIEW™; (2) the RBEC reactor unit; (3) reference electrode; (4)
variable external resistor; (5) central rotating shaft; (6) electrode disc assembly; (7)
computer controllable stepper motor; (8) a computer controllable relay switch interfacing
the half-discs and the potentiostat; (9) peristaltic pump; (10) heat exchanger to warm up
the cold feed; (11) influent feed reservoir, at ~4°C ice bath; (12) effluent holding tank, at
~4°C ice bath. Note: figure not drawn to scale; RE: reference electrode; CE: counter
electrode (i.e. cathode); WE: working electrode (i.e. anode).
A rotatable electrode disc assembly consists of a Perspex horizontal central rotatable
shaft. Two separate stainless steel current collecting strips were independently mounted onto the
Potentiostat
RE CE WE
Process
Monitoring and Control
(1)(8)
(9) (9)
(11)
(12)(7)
(2) (3)
(10)
(6)
Open Air
Headspace
(5)
(A) (B)
e - e -
V
Process
Monitoring
and Control
Open Air Headspace
Ω
(1)
(2)
(6) (7)
(4)(3)
(5)
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 119 -
upper and lower sides of the shaft. 20 sets of electric conductive discs were mounted onto a
central horizontal rotatable shaft.
Photo 7.1 Photographs of the bench-scale RBEC reactor. (A) A stepper motor is used to rotate
the electrode-disc assembles in the two separate reactors; (B) a side-view of a single reactor; (C)
water level reaches the bottom edge of the upper cathodic half-disc to complete ionic circuit.
A computer controllable stepper motor unit consists of a stepper motor, a stepper motor
driver (K179 Stepper Driver), and an analog output card (Labjack U12). The motor was tightly
mounted onto the side Photo of the reactor vessel. A gear disc (one-to-seven reduction) was used
to convey the mechanical action of the motor to the central rotatable shaft.
7.2.2 Design of Two Half Discs Serving as Anode and Cathode
Each disc consists of two halves; an upper air-exposed and a lower water submerged half
(Figure 7.2, Photo 7.2). Each half consists of a stainless steel current collecting mesh.
Electrically conductive carbon fiber sheet was mounted onto both sides of the mesh. The upper
and the lower halves were physically and hence electrically separated from each other. Each half
has a small extension of the stainless steel mesh which serves as the electric contact with a
stainless steel current collector located at the central shaft. Electrical connection between the
stainless steel current collecting strips and the external circuit was completed with two
individual copper electrical wires. A Perspex plastic o-ring was used to connect each pair of
electrode discs with the central shaft and to maintain a constant distance with the neighboring
discs. The estimated total projected surface area (water submerged) of the 20 disc halves was
947 cm2. This value is used for calculating the current and power densities. After filling up the
1
2
A Stepper motor
C B
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 120 -
RBEC reactor with the wastewater, about 5 mm of the bottom edge of the air-exposed half discs
were immersed in the water. This is essential to allow an ionic contact between the air- cathode
and the submerged anode such that the system does not require the use of a membrane or
separator.
Figure 7.2 A schematic diagram of an electrode disc set. (1) A Perspex plastic o-ring; (2) a
small extension of the stainless steel mesh. Note: figure not drawn to scale.
Photo 7.2 Photographs of the rotatable electrode-disc assemble. (A) Two electrode half-discs are
mounted on a perplex O-ring; (B) stainless steel mesh enables electrical contact between
the carbon-sheet electrode and the central shaft where a stainless strip is mounted thereon;
40 mm
40 mm
120 mm
8 mm5 mm
Water level
Water Submerged
Air-Exposed
Carbon sheet
mounted on stainless steel mesh
(1)
(2)
1
2
3
Stainless-steel current collector
Copper wires external circuit
C A B
D
Rotating Shaft
E
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 121 -
(C) 20 electrode-disc assembles are mounted onto a central shaft; (D) two separate
external copper electrical wires are used to convey electrons to or from the upper and
lower discs respectively; (E) stainless steel mesh serves as a support and current collector
of the carbon sheet.
7.2.3 Process Monitoring and Control
In general, a computer program LabVIEW™ (version 7.1 National Instruments™) was
developed to continuously control and monitor the RBEC process (see Photo 7.3). Voltage
signals were recorded at fixed time intervals via the LabVIEW program interfaced with a high
precision voltage data acquisition board (DAQ) (National Instruments™ NI4350). All electrode
potential (mV) described in this study refers to values against Ag/AgCl reference electrode (ca.
+197 mV vs. standard hydrogen electrode).
Photo 7.3 Computer programs were written to control and monitor the RBEC bioprocess: (A)
RBEC operated as a MFC mode for net electricity generation; (B) RBEC coupled with a
potentiostat.
When the RBEC was operated as a MFC, the potential differences between the air-
exposed cathode and water submerged anode (i.e. cell voltage); and between the submerged
discs and a silver-silver chloride reference electrode (Bioanalytical System, RE5) were
continuously monitored by the DAQ. When a potentiostat was used to supply extra electrical
power to the system and to maintain a constant potential of the water submerged anode, the
current was directly recorded from the potentiostat.
A B
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 122 -
Rotation of the electrode discs was fully automated. The stepper motor and its driver is
calibrated such that upon receiving a proper signal from the computer program (via a Labjack™
analog output card), the electrode discs could be rotated at a defined time interval either for a
full-(360°) or a half-(180°) rotation (intermittent-flipping). In the latter case, each half disc could
intermittently serve as a submerged anode and an air-exposed cathode (Figure 7.3). When the
RBEC was coupled to a potentiostat, a computer controllable relay-switch was fabricated to
intermittently switch over the connection between the working electrode of the potentiostat and
the two half discs (see Photo 7.4). The air-exposed halves always serve as the counter electrode.
Hence, both half discs would alternately expose to an identical anodic and cathodic environment
during the intermittent-flipping operation.
Figure 7.3 Sequential disc rotation and control of the submerged anode potential during an
alternating flipping operation of the RBEC reactor.
Photo 7.4 A computer controllable relay-switch for intermittently switching over the connection
between the working electrode of the potentiostat and the two half discs.
Air
Water2..
1.Air
Water
Cathode
Air
Water1.
2.Air
Water
1st
2nd
3rd
4th
180 Clockwise
Rotation
180 Anti- Clockwise Rotation
Anode
Anode
Cathode
1.
2.
2.
1.
Connect to
RBEC Connect to
Potentiostat
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 123 -
7.2.4 Bacterial Inoculum and Synthetic Wastewater
A return activated sludge collected from a local municipal wastewater treatment plant
was used as the inoculum. It was stored at 4°C prior use. A synthetic wastewater was used
throughout the study. Unless otherwise stated, it consisted of (mg L-1
): NH4Cl 125, NaHCO3 125,
MgSO4·7H2O 51, CaCl2·2H2O 300, FeSO4·7H2O 6.25, and 1.25 mL L-1
of trace element
solution, which contained (g L-1
): ethylene-diamine tetra-acetic acid (EDTA) 15, ZnSO4·7H2O
0.43, CoCl2·6H2O 0.24, MnCl2·4H2O 0.99, CuSO4·5H2O 0.25, NaMoO4·2H2O 0.22,
NiCl2·6H2O 0.19, NaSeO4·10H2O 0.21, H3BO4 0.014, and NaWO4·2H2O 0.050. Sodium acetate
(5 to 10 mM) was the sole electron donor in this study. Yeast extract (0.1-1 g L-1
final
concentration) was added as bacterial growth supplement. Unless otherwise stated, the initial pH
of the synthetic wastewater was adjusted to 6.9-7.2 using either 1M HCl or 4M NaOH. No
additional pH buffer was added.
7.2.5 Reactor Operation
7.2.5.1 Start-up as a Microbial Fuel Cell
The RBEC was loaded with an activated sludge amended (10%, v/v) synthetic
wastewater (1.75 L). It was operated in batch mode as a MFC for about two months (Figure
7.1A). A variable external resistor (2 - 1M Ω) was used to connect the upper and the lower disc
half, completing the electric circuit of the system. Unless otherwise stated, the RBEC was
operated at ambient temperature (22 -25oC) and atmospheric pressure. Polarization curve
analysis was regularly conducted to evaluate MFC performance over time (Logan et al. 2006).
Briefly, it was done by decreasing the external resistance from 1M to 5 Ω in a stepwise manner.
For each external resistance setting, at least 5 min waiting period was given to obtain steady
state values of current (I) and voltage (V). These values were used to obtain the corresponding
power output (P) according to an equation of P = V × I. Power and current densities were
obtained by normalizing the P and I with the projected surface area of the submerged half discs
(i.e. 947 cm2), respectively. Correlation between the submerged anode potential and the current
density was further obtained from the data of the polarization curve analysis. This correlation is
useful to evaluate anodophilic property of the electrochemically active biofilm in a MFC (Cheng
et al. 2008). Acetate concentration of the wastewater was regularly quantified to obtain COD
removal rate at different runs (Cheng et al. 2008). Chemical oxygen demand (COD) was derived
from acetate concentration assuming 1.067 g COD·g acetate-1
.
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 124 -
After start up, COD removal rates at external resistances of 2 Ω and open circuit were
compared to test if allowing current generation in the RBEC could accelerate COD removal. The
effect of disc rotation (full-turn, half-turn or no rotation) on COD removal was also assessed at
open circuit.
7.2.5.2 Coupling the RBEC with a Power Source
After a two months operation as a MFC, the RBEC was coupled with an electrical power
source (potentiostat) to overcome any cathodic limitation in the system. The working (WE),
counter (CE) and reference electrodes (RE) of the potentiostat (Model no.362, EG&G, Princeton
Applied Research, Instruments Pty. Ltd.) were connected to the submerged anode, air-exposed
cathode and the reference electrode, respectively (Figure 7.1B; Photo 7.5). Unless otherwise
stated, the anode potential was controlled at -300 mV.
The potentiostat-assisted RBEC was initially operated in fed-batch mode for about a
week before switching into continuous mode. A synthetic wastewater (10 mM acetate) was
continuously fed through the reactor at a flow rate between 2.5 and 6 ml min-1
, corresponding to
a hydraulic retention time (HRT) between 4.9 and 11.7 h; and an organic loading rate (OLR)
between 1.3 and 3.1 kg COD·m-3
·d-1
. COD removal rate was obtained according to: (CODIn
-
CODOut
)/HRT, where CODIn
and CODOut
are COD concentrations (mg·L-1
) of the influent and
effluent, respectively. Both the influent reservoir and the effluent were stored at 4oC to avoid
microbial deterioration.
7.2.6 Scanning Electron Microscopy of Biofilm-Electrode
The morphology of the established electrode biofilm was examined by using scanning
electron microscopy (SEM) (Philips XL 20 Scanning Electron Microscope). Small pieces of
carbon electrode were removed from the reactor with a sterile stainless steel scissor. The
electrode samples were fixed in 3 % glutaraldehyde in 0.025 M phosphate buffer (pH 7.0),
overnight at 4 oC. The samples were then rinsed three times (5 min each time) with 0.025 M
phosphate buffer (pH 7.0) before they were subjected to dehydration procedure through a series
of ethanol solution (each solution was changed two times and each change was 15 min): 30, 50,
70, 90 and 100%. The 100% ethanol solution was replaced with amyl acetate (two changes, 15
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 125 -
min each. This step was performed in fume hood). The samples were critical point dried before
they were mounted to SEM specimen holder with contact adhesive. The samples were then
sputter coated with gold before being subjected to SEM observation. A fresh carbon electrode
was used as the abiotic control.
Photo 7.5 Chemostat operation of the RBEC reactor coupled with an external power supply
(potentiostat). (1) RBEC reactor; (2) potentiostat; (3) influent feed; (4) effluent tank; (5)
ice bath; (6) peristaltic pump; (7) computer for process control and monitoring.
(1)
(2)
(3) (4)
(5)
(6)
(7)
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 126 -
7.3 Results and Discussion
7.3.1 Operation as a Microbial Fuel Cell for Electricity Generation
The described reactor was constructed, inoculated with activated sludge and operated as
a fed-batch MFC reactor for about two months.
Unlike other membrane equipped, two-chamber MFC where an anode is usually
maintained under strict anaerobic condition, the absence of a membrane in the RBEC would
inevitably allow oxygen to diffuse from the air-headspace to the anodic region in the water
phase. Yet, a reasonably reducing redox state of this anodic region is essential for electricity
generation. In essence, the anode must first be polarized (bio-electrochemically) at a potential
low enough to drive electron flow towards the cathode. This could be achieved by the addition
of acetate (ca. 10 mM), amended activated sludge (10%, v/v) to the RBEC, which reduced the
anodic potential to a steady level of -453 mV at 1M Ω within 16 hours (Figure 7.4).
Figure 7.4 Effect of activated sludge on the evolution of anode potential and cell voltage of the
RBEC reactor during the initial startup. External resistance was 1M ohm. Activated
sludge was inoculated at time zero.
7.3.1.1 Increased Anodophilic Activity increases Power Output over Time
The RBEC was operated as explained under Section 7.2 by rotating full turns every 15
min. This allowed the activated sludge bacteria to colonize both, the submersed anode half and
-600
-400
-200
0
200
400
600
0 5 15 20Time (h)
Vo
lta
ge
(m
V)/
Po
ten
tia
l (m
V v
s. A
g/A
gC
l) Voltage-10% AS
Voltage-0% AS
AP-10% ASAP-0% AS
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 127 -
the air exposed cathode half of each disc. Electricity generation of the RBEC was evaluated over
50 days by using polarization curve analysis. Similar to other MFCs, the current and power
outputs of the RBEC reactor increased over time (Figure 7.5) reaching a maximum current
density and power density of 0.075 A·m-2
and 8.4 mW·m-2
respectively.
Figure 7.5 Polarization and power density curves recorded at Day 1, 14, 38 and 50. Reactor was
operated in Batch mode (acetate concentrate ranged from 5 to 11 mM).
The observed increase of current output could in theory be due to microbial enhancement
of either the cathodic or anodic half-reaction. To determine which of the two half-reactions was
stimulated by the developing biofilm, a series of current/potential curves was plotted (Figure
7.6). A clear increase in the anodic reaction could be observed from the fact that the bacteria
could generate more current for the same anodic potential. For example, at an anodic potential of
-500 mV the bacteria could generate twice as much current on day 50 compared to day 38
0
100
200
300
400
500
600
Vo
lta
ge
(m
V) Day 1
Day 14Day 38Day 50
0
1
2
3
4
5
6
7
8
9
0 0.02 0.04 0.06 0.08
Current Density (A m )-2
Po
we
r D
en
sity (m
W m
-2)
Day 1Day 14Day 38Day 50
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 128 -
(Figure 7.6). This indicates that over time the biofilm has become more capable of using the
water-immersed anode as their electron acceptor.
Figure 7.6 Current density as a function of anodic potential of the RBEC reactor obtained at
different times. The reactor was operated at batch mode with the medium saturated with
acetate.
7.3.2 In Situ Supply of Oxidizing Power via Conductive Contactor (as Electron Flow)
increases COD Removal
Although the design of the RBEC is similar to a RBC, their underlying principles of
COD removal are different. In RBC, the contactor discs are continuously rotated (1-10
revolutions per minute) to provide oxidizing power (dissolved oxygen) for the biofilm to
aerobically degrade the COD (Cortez et al. 2008; Di Palma and Verdone 2009). COD removal
can be accelerated by rotating the contactor faster to increase mass transfer of dissolved oxygen
to the biofilm (Cortez et al. 2008; Israni et al. 2002). A concentration gradient of both the COD
(electron donor) and oxygen (electron acceptor) exists along the biofilm depth, of which the
outer layer (near the bulk) always has a higher concentration compare to the inner layer (near the
contactor). While in RBEC, oxidizing power is provided in-situ from the contactor surface. The
biofilm can oxidize the wastewater COD by using the contactor directly as their electron
acceptor, alleviating the demand of disc rotation for oxygen mass transfer.
0
0.02
0.04
0.06
0.08
-600 -500 -400 -300 -200 -100
Anodic Potential (mV vs. Ag/AgCl)
Cu
rre
nt D
en
sity (A
m -2) Day 1
Day 14Day 38Day 50
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 129 -
Figure 7.7 Changes in (A) current, (B) anodic potential and (C) acetate concentration in the
RBEC reactor operated at closed or open circuit.
To test whether by allowing electron flow from the water-immersed to the air-exposed
half disc could accelerate COD removal, COD removal rates in the RBEC operated at closed (2
Ω) or open circuit were compared. The discs were regularly rotated (1 turn at 1rpm every 15 min)
to approach RBC operation (Figure 7.7). At open circuit, the regular disc rotation could lead to a
52% increase in COD removal rate (from 0.32 to 0.48 kg COD·m-3
·d-1
) (Figure 7.7C). Similar to
the conventional RBC, the rotating action of the discs would have increased the oxygen mass
transfer and resulted in a faster microbial COD degradation. Current generation could further
increase COD removal rate by 35% (from 0.48 to 0.65 kg COD·m-3
·d-1
) (Figure 7.7C). Hence,
allowing electron flow from the water-immersed to the air-exposed half disc could accelerate
COD removal. This implicates that the treatment performance of traditional RBC may be
significantly improved by simply using conductive contactor materials.
0
1
2
3
4
5
6
Cu
rre
nt(
mA
) .
-550
-450
-350
-250
-150
-50
An
od
ic P
ote
ntial.
(mV
vs. A
g/A
gC
l)
. 2 Ohm
Open Circuit
0.43 mM·h-1
(R2 = 0.99)0.21 mM·h-1
0
1
2
3
4
5
0 5 10 20 25 30Time (h)
Ace
tate
Co
ncentr
atio
n
(mM
) .
2 Ohm (Rotated 360°Every 15 min)Open Circuit (Rotated 360°Every 15 min)Open Circuit (No Rotation)
0.32 mM·h-1
(R 2 = 0.99)
(R 2 = 0.99)
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 130 -
7.3.2.1 Parasitic Current decreases Coulombic Recovery
Although allowing current generation would favor COD removal in the RBEC, only low
Coulombic recovery was obtained (4.2%). This would most likely be due to a so-called
―parasitic current‖ effect as recently described by Harnisch and Schroder (2009). Since the
cathode discs were partially (~8% of the projected surface area of the half disc) immersed in the
wastewater to maintain ionic balance in the absence of an ion-exchange membrane, the bacteria
at the submerged region of the cathode might anodically oxidize the COD in the bulk via an
internal short circuit reaction at the cathode. As about 92% of the projected surface area of the
cathode was exposed to the air, the electrons generated at the submerged cathodic region would
readily flow towards the air-exposed counterpart and finally react with oxygen (Figure 7.8). This
short circuit current would by-pass the external circuit of the system and became unaccountable
for the Coulombic estimation. Further, the anodic reaction at the cathode would ―drag down‖ the
apparent cathodic potential to a more negative level (mixed potential). As a consequence, the
open cell voltage is diminished.
Figure 7.8 A schematic illustration of the principle of the parasitic current that could occur at
the air-exposed cathodic half disc of the RBEC. Note: diagram is not drawn to scale.
Modified after Harnisch and Schroder (2009).
Ext. Ω
Biofilm
Water
COD
CO2 + H+
O2 + 4H+
H2O
e-
e-
e-
e-
COD
O2
Parasitic Current
Anode
CathodeE°’
O2/H20
CH2O/CO2
EmixedAir
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 131 -
To evaluate the effect of parasitic current, the RBEC was operated at two different water
levels: 1) Low level where the entire cathode discs were exposed to the air. Ionic flux was
maintained by placing a non-conductive polymer sponge (as a salt-bridge) between the two
electrode halves; 2) High level where the cathode discs were partially immersed in the water as
in the original setting. The cell voltage and electrode potentials were recorded over time (Figure
7.9).
At low water level, the cell voltage was about 400 mV. Increased the water level resulted
in a gradual decrease in cell voltage to about 270 mV. Such decline was not due to the anodic
but to the cathodic half reaction, which had gradually decreased the cathodic potential from -180
to a more negative, mixed potential of -330 mV (Figure 7.9). This signifies the presence of a
parasitic current at the cathode, which represents an alternative path of COD removal even at
open circuit operation (Figure 7.10). From an electricity generation standpoint, the parasitic
current is undesirable as it diminishes the Coulombic recovery. However, it would still be
desirable from a wastewater treatment standpoint that the bacteria can degrade COD using the
conductive contactors of the RBEC at either close or open circuit.
Figure 7.9 Effect of water level on the electron potentials and cell voltage of the RBEC under
open circuit. Notes: A non-conductive sponge was sandwiched between the anodic and
cathodic discs to allow ionic flux during low water level operation; reactor operated at
batch-mode with acetate saturation (>2mM). pH 8.5 at 35oC; the discs were rotated for a
full turn once per hour at 1 rpm.
-600
-400
-200
0
200
400
0 5 10 15 20 25 30Time (h)
Cell
Voltage (
mV
)/
Ele
ctr
ode P
ote
ntial (m
V v
s. A
g/A
gC
l)
Cell Voltage
Cathode
Anode
Open Circuit
Cathode
Anode
Low Cathode
Anode
High
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 132 -
Figure 7.10 Effect of external current on (A) acetate removal and (B) electrode potentials.
RBEC operated at batch mode with the cathodic half discs partially immersed in the
water, pH 8.5 at 35oC; the discs were rotated for a full turn once per hour at 1 rpm.
7.3.3 Sequential Flipping the Electrode Discs allows Alternate Current Generation
Chapter 6 suggests that a BES biofilm could catalyze both anodic substrate oxidation and
cathodic oxygen reduction in a two-chamber BES, causing an intermittent polarity inversion of
the system. Such bidirectional microbial electron transfer could alleviate both the pH gradient
and cathodic oxygen reduction limitations in the BES (Chapter 6). The same concept was tested
here with the RBEC. The discs were intermittently flipped (180°, every 1h) instead of fully
rotated such that each half disc was alternately exposed to the wastewater and the air (see Figure
7.3). The result showed that the intermittent-flipping action could revert the polarity of current
(ranged from -6.7 to +7.5 mA), corroborating with the finding that the biofilm could alternately
catalyze the current under alternating redox conditions (Figure 7.11A). An averaged power
density of 0.93 mW·m-2
was obtained. In terms of COD removal, allowing current generation
under the intermittent-flipping operation could accelerate the COD removal rate by 15%
compared to the open-circuit control (from 0.79 to 0.91 kg COD·m-3
·d-1
) (Figure 7.11B). This
-600
-500
-400
-300
0 5 10 15 20 25 30Time (h)
Po
ten
tia
l (m
V v
s. A
g/A
gC
l)
0
5
10
15
20
Cu
rre
nt (
mA
)
0
1
2
3
4
5
Ace
tate
(mM
)
Open circuit Close circuit (2Ω)
0.20 mM∙h-1
0.11 mM∙h-1
Cathode
Anode
Acetate
(B)
(A)
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 133 -
again suggests that allowing electron flow from the water-immersed to the air-exposed half disc
could accelerate COD removal.
Figure 7.11 Effect of regular half-rotation of electrode dices on (A) current, anodic potential and
(B) acetate concentration in the RBEC reactor operated at close (2 ohm) and open circuit.
The discs were flipped (180°) every 1 hour. Solid arrows in A indicate flipping events
(~1min).
Interestingly, when current was prohibited (open circuit) the intermittent-flipping
resulted in a 2.5-fold increase in COD removal rate compared to the control without rotation
(from 0.32 to 0.79 kg COD·m-3
·d-1
). Since the biofilm was alternately exposed to the anoxic
wastewater and oxygen, under oxygen limited condition (i.e. submerged in the wastewater) the
biofilm may undertake storage of substrate (acetate) from the wastewater as intracellular
polymers such as polyhydroxybutyrate (PHB) (Van Aalst-van Leeuwen et al. 1997; Van
Loosdrecht et al. 1997). Subsequent exposure to the air may allow the bacteria to oxidize the
storage materials with oxygen. Hence, apart from the parasitic current effect, bacterial storage
-10
-5
0
5
10
Cu
rre
nt (m
A)
0.60 mM·h-1
0.52 mM·h-1
(R2
= 0.99)
0
2
4
6
8
10
0 5 10 20 25 30 35Time (h)
Ace
tate
Co
nce
ntr
atio
n
(mM
)
Open Circuit
2 Ohm
Half-Turn (180°) Disc Rotation Every 1h:
(R2
= 0.95)
(A)
(B)
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 134 -
may account for the enhanced COD removal under the intermittent-flipping operation of the
RBEC. Yet, further study is required to clarify this point.
7.3.4 Coupling the RBEC with an External Power Source to achieve Higher Current
Similar to most other air-cathode based MFC, the sluggish cathodic oxygen reduction has
also limited the electricity generation in the RBEC (Clauwaert et al. 2009; Kim et al. 2007;
Wang and Han 2008). As a result, the anodic half discs became an unfavorable electron acceptor
for the biofilm (i.e. very negative potential) even when the external resistance was not limiting.
In the literature, cathodic limitation in a BES could be overcome by providing an extra voltage
to ―boost‖ the cathodic reaction. A well known example is cathodic hydrogen production using
MEC (Logan et al. 2008). Here, a potentiostat was used to overcome the cathodic overpotential
and to keep a constant, more suitable (less negative) anode potential (-300 mV) for the biofilm
to ―respire‖. However, instead of generating hydrogen under anaerobic condition the cathode
remained as oxygen based.
7.3.4.1 Electrochemically Assisted Anode Facilitates the Establishment of Anodophilic
Biofilm
The RBEC was coupled to a potentiostat as described in the method section. In the
absence of acetate, a background current of ~ 2 mA and a cathodic potential of -300 mV were
obtained (Figure 7.12A, B). The acetate addition resulted in a similar current profile as observed
in the previous, non-potentiostatic assisted mode but at about 10-fold higher current (about 60
mA) (Figure 7.12A). The increased current was caused by the potentiostat, which has provided
the additional potential to overcome the cathodic reduction limitation. In this example where the
anode was maintained at -300 mV, a cathodic potential of about-1800 mV was required to
substantiate the current triggered by the anodic substrate oxidation (Figure 7.12B). A higher
Coulombic recovery of 40% (versus ca. 5% in non-assisted mode) was obtained, indicating that
the biofilm has become more anodophilic with the more ―attractive‖ anode.
After a several fed-batch cycles, the RBEC was switched into a continuous mode to
avoid substrate limitation (Figure 7.12C, D). The current was maintained at a similar level
between 40 and 60 mA as observed during the batch-mode operation. The average COD
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 135 -
removal rate was 1.15 ± 0.05 kg COD·m-3
·d-1
. Overall, the anodic biofilm has become more
anodophilic with a more suitable anodic potential.
Figure 7.12 Performance of a potentiostat-controlled RBEC operated at (A, B) batch mode and
(C, D) continuous mode. The submerged anode discs were poised at -300 mV vs.
Ag/AgCl. The discs were rotated (360°) once per 15 min.
Figure 7.13 Current-potential plots suggest that the two half discs have (A) similar cathodic
activity, but (B) different anodophilic activity. WE: working electrode; CE: counter
electrode.
0
20
40
60
80
Cu
rre
nt
(mA
)
48 Time (h) 84 9672
Anode
Cathode
1.10 kg COD·m-3·d-1
1.19 kg COD·m-3·d-1
1.16 kg COD·m-3·d-1
Ac
eta
te S
tarv
ed
(Ba
tch
mo
de
)
Acetate Saturated
(Continuous mode)
0
20
40
60
Cu
rre
nt
(mA
)
1
2
3
4
5
6A
ceta
te (
mM
)
-1800
-1400
-1000
-600
-200
0 12 20 3624
Po
ten
tia
l (m
V v
s.
Ag
/Ag
Cl)
Anode
Cathode
-1800
-1400
-1000
-600
-200
Acetate
(A) (C)
(B) (D)
0
20
40
60
-800 -400 0 400 800
Working Electrode (WE) Potential (mV vs. Ag/AgCl)
-60
-40
-20
0
-1500-1000-5000
Cu
rre
nt (m
A)
Half discs mostly exposed to air Half discs mostly submerged in water
(A): (B):WE
CE
CE
WE
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 136 -
Prolonged operation with periodic full disc rotation resulted in a difference in the
bioelectrochemical properties between the two half discs as each half would serve
predominantly as either an anode or a cathode. The relationship between the current and
electrode potential indicates that the two half discs exhibited similar cathodic but different
anodic properties (Figure 7.13).
7.3.4.2 Sequential Flipping the Electrochemically-Assisted Discs establishes
Anodophilic Biofilm on both Half Discs
In order to establish a suitable biofilm capable of catalyzing both an anodic acetate
oxidation and cathodic oxygen reduction on both half discs, the process was modified such that
the disc was only rotated for half a turn (180 degree) instead of a full turn (360 degree). Hence,
each half disc could serve as a submerged anode and an air-exposed cathode intermittently. The
submerged anodic half discs were always controlled at fixed potential of -400 mV (Figure 7.14).
Figure 7.14 Intermittent Disc-flipping operation of the RBEC with the submerged anodic discs
controlled at a potential of -400 mV. (A) Anodic current over time; (B) electrode disc
potential over time; Prior to the experiment, the biofilm at disc 2 was more anodophilic
than that at disc 1 (Figure 7.13). In the first 16 h the discs were flipped once every 30
min, thereafter they were flipped once per hour.
0
10
20
30
40
50
Cu
rre
nt (
mA
)
-1200
-1000
-800
-600
-400
0 24 48 72 96Time (h)
Po
ten
tia
l (m
V v
s. A
g/A
gC
l)
Fixed Anodic Potential
1
2
2
1
Intermittent Flipping cathode
anode
cathode
anode
Cathodic Potential
(A)
(B)
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 137 -
Initially, anodic current was not observed when disc 1 (the half discs mostly exposed to
the air in previous runs) served as the anode because of the lack of anodophilic activity as
suggested in Figure 7.13. Flipping over the discs to revert the polarity could enable anodic
current of about 30 mA. This current was catalyzed by the anodophilic biofilm at disc 2 (the half
discs mostly submerged in water in previous runs) (Figure 7.14A).
Operating the RBEC for a longer period under the intermittent-flipping regime gradually
increased the anodic current catalyzed by the biofilm at disc 1 (Figure 7.14A). This current
improvement also coincided with an increase in COD removal rate over time. At 51, 61 and 92 h
the COD removal rates were 0.60, 0.88 and 1.31 kg COD·m-3
·d-1
, respectively. Therefore, the
intermittent flipping of disc and polarity inversion could establish electrochemically active
biofilm at both half discs, contributing to a higher COD removal from the wastewater.
7.3.5 Intermittent Flipping of the Discs avoids the Continuous Alkalization of the
Cathode
The cathodic oxygen reduction consumes protons according to equation: O2 + 4e- + 4H
+
→ 2H2O. Since the protons and hydroxide ions are always in equilibrium with each other
through the water dissociation (Kw=[H+][OH
-] ≈ 10
-14), net consumption of proton represents a
hydroxide ion production: O2 + 2H2O + 4e- → 4OH
-. Hence, it is expected that the localized pH
at the cathode disc would increase over time. According to Nernst equation, a 10 fold increase of
hydroxide ion (i.e. ΔpH = +1) causes a reduction of cathodic potential by 59 mV. This
represents a loss of electromotive force (e.m.f) in the system and hence more energy is required
to sustain the forward cathodic reaction. In the electrochemically-assisted RBEC, this extra
energy is compensated by the external power source (potentiostat). Thus, neutralizing the
cathode alkalinity could reduce the electricity input in the process.
The effect of intermittent flipping of the electrode discs on the cathode pH was evaluated
(Figure 7.15). Before a flipping event (from 5 to 65 min), the cathode pH (disc 1) increased from
7.0 to 9.0 and the cathode potential became more negative (-74 mV/pH). A similar trend was
observed for another half disc (disc 2) from 65 to 125 min. During this period, disc 1 was
immersed in the wastewater (pH about 7) serving as an anode and its previously established
alkalinity was neutralized. After the next flipping (125 min), the cathode pH and potential
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 138 -
resumed to their original levels (pH 7 and -1270 mV, respectively). The continuous build-up of
alkalinity at the air-cathode could thus be avoided.
Figure 7.15 Effect of intermittent flipping of the electrode discs on the pH of the air-exposed
cathode. The discs were flipped at about 5, 65 and 125 min.
7.3.6 The Established Biofilm could catalyze a Cathodic Oxygen Reduction
After the biofilm was acclimatized in a RBEC with intermittent exposure to the air and
with alternating current generation, the catalytic property of the established biofilm electrode for
cathodic oxygen reduction was compared with a control (abiotic) electrode using voltammetric
technique (LSV) (Figure 7.16). In a potential range between -0.2 and -1.0 V, the biofilm
electrode outperformed the control electrode with respect to current generation. At -0.5 V the
current density of the biofilm electrode was -0.24 A m-2
, which was about 2 fold higher than that
of the abiotic control (-0.12 A m-2
). However, the cathodic overpotential of oxygen reduction
remained significantly large, and the presence of biofilm did not appear to be beneficial on
alleviating the overpotential of the cathodic reaction. This result is in contrary to the previous
7
8
9
Ca
tho
de
pH
0
50
100
150
200
250 Cu
rren
t (mA
)
Cathode pH Current
-2000
-1500
-1000
-500
0
0 30 90 120Time (min)
Po
ten
tia
l (m
V v
s. A
g/A
gC
l) .
Submerged Discs (Anode)
Air-Exposed Discs (Cathode)
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 139 -
chapter where an anodophilic biofilm could significantly reduce the overpotential of cathodic
oxygen reduction at a granular graphite electrode.
Figure 7.16 Effect of biofilm on the cathodic oxygen reduction. Potential scan rate 1.0 mVs-1
,
Current density was calculated based on a projected surface of the carbon electrode of 7
cm2, dissolved oxygen concentration in the medium was maintained at >6 mg/L.
The morphology of the same biofilm electrode was examined with SEM technique
(Figure 7.17). Compared to the abiotic control (A), a developed biofilm was established on the
electrode fibers. However, a boarder examination of the samples reveals that not all of the fiber
surfaces were colonized by the biofilm (Figure 7.17C).
Figure 7.17 Scanning electron micrographs of (A) a plain carbon fiber sheet; and (B to D)
carbon fiber sheet attached with the biofilm obtained from the RBEC under alternate disc
flipping operation (aerobic headspace).
1.0
1.2
1.4
1.6
1.8
2.0
-1-0.8-0.6-0.4-0.20
Cathodic Potential (mV vs. Ag/AgCl)
Cu
rre
nt
De
nsity (
A m
-2)
-0.7
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
Ra
tio o
f Bio
/ Ab
iotic
Cu
rren
t
Ratio Bio/Abio
Biofilm Electrode
Abiotic Electrode
1
2
3
A B
C D
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 140 -
7.3.7 Energy Evaluation of the Electrochemically Assisted RBEC process
The treatment performance and energy requirement of the process are compared with
that of conventional AS processes (Table 1). The COD removal rate achieved by the RBEC fall
within the range of the conventional AS processes, suggesting that the RBEC process is equally
good as compared to the AS processes in removing organic pollutants from the wastewater. In
terms of energy usage, the RBEC process demonstrated better efficiency (0.47 kWh kg COD-1
)
compared to the AS processes (0.7-2.0 kWh kg COD-1
). Further, the RBEC process also
demonstrated at least 5 times less energy demand per volume of wastewater treated as compared
to the AS processes (88 vs. 430 kWh ML-1
) (Table 1). These figures indicate that a substantial
energy saving may be achieved if the wastewater COD was degraded via electric conductor as
electron flow instead of via forced-aeration of wastewater as in the conventional AS processes.
Table 7.1 Treatment performance and energy requirement of the RBEC process and
conventional activated sludge (AS) processes.
Conventional AS
Processes RBEC Process
a
#COD removal rate (kg COD m
-3 day
-1) 0.5 - 2
b 1.32
Energy input per COD removed (kWh kg COD-1
) 0.7 - 2 c
¥0.47
Energy per volume of wastewater treated (kWh ML-1
) 430 - 940 d
¥88
a Values for the RBEC process are calculated based on data obtained in Figure 7.14 (91-100 h).
b Adapted from (Logan et al. 2006)
c Adapted from (Tchobanoglous et al. 2003)
d Adapted from (Keighery 2004).
#COD content of acetate was estimated assuming 1.067 g COD per 1.0 g acetate.
¥The values include the energy consumption by the stepper motor (0.083 Wh per one rotation).
7.4 Concluding Remarks
Overall, this chapter has demonstrated the working principle of the proposed RBEC
process as a BES. Based on the findings, the following points are summarized:
The RBEC can operate as a MFC, achieving net electricity recovery while removing organic
pollutants from the wastewater. Even without adjusting the pH of the wastewater, the
detrimental problem of pH split as observed in membrane-BES was not encountered.
Chapter 7: Development of a Rotatable Bio∙Electrochemical Contactor
- 141 -
Intermittent disc flipping and polarity inversion establishes anodophilic/ cathodophilic
biofilm at both half discs, leading to increasing pollutant removal rate.
The RBEC bioconversion rate could be increased by controlling the anodic potential with an
external power source (potentiostat).
Intermittent flipping of the discs avoids the continuous alkalization of the cathode.
The biofilm could catalyze a cathodic oxygen reduction. However, the cathodic
overpotential of oxygen reduction remained significantly large.
The electrochemically assisted RBEC process can remove COD more energy efficient than
the conventional activated sludge process.
Treatment performance of traditional RBC process may be significantly improved by simply
using conductive contactor materials.
Further process optimization is required to maximize the functionality of the RBEC to
recover useful energy from wastewaters. These efforts may include: (i) optimizing current
density and COD removal by increasing the electrode surface area (e.g. number of contactor
discs) per liquid volume ratio; (ii) minimizing overpotential of the cathodic oxygen reduction by
modifying the electrode with a suitable catalyst or using more effective electrode materials; (iii)
apart from the direct production of electricity, the possibility of the RBEC for the production of
other valuable products or energy carriers such as hydrogen or methane gas should also be
explored.
- 142 -
8 A Rotatable Bioelectrochemical Contactor (RBEC)
enables Electrochemically Driven Methanogenesis
Chapter Summary
In this chapter, the RBEC prototype was operated as a microbial electrolysis cell (MEC)
under fully anoxic condition. The capacity of the system to couple anodic acetate oxidation with
a cathodic production of energy carrier gases (e.g. hydrogen or methane) was evaluated.
When the oxygen (air) in the headspace was replaced with nitrogen, hydrogen was
produced. However, prolonged anoxic operation resulted in the production of methane (≥50%,
v/v) instead of hydrogen (~0.1%). Such methane production was directly proportional to both
the current and the electrical energy input. For example, increased the current from 15 to 390
mA (0.16 and 4.1 A m-2
) resulted in a 7-fold increase in the COD removal rates (from 0.2 to
1.38 kg COD m-3
day-1
) and a 13-fold increase in the methane production rate (from 0.04 to 0.53
L L-1
day-1
). Over 80% of current was recovered as methane. Using the described RBEC to
convert COD (here acetate) into methane could allow energy saving and possible energy
recovery from the wastewater (-1.5 to 1.5 kWh kg COD-1
).
Although the proposed method of methane production is unlikely to replace traditional
anaerobic digestion, the fact that the bioconversion and methane generation rates could be
controlled electrochemically appears promising. Further, the absence of precious chemical
catalysts (e.g. platinum), ion exchange membranes and the low requirement of liquid
recirculation may represent reduction in capital cost, facilitating process up-scaling.
Overall, the RBEC described in this chapter may represent an alternative option for
large-scale BES-based technology for energy recovery from wastewater. The use of RBEC in
other application niches should also be explored in future.
Chapter 8: RBEC enables Electrochemically Driven Methanogenesis
- 143 -
8.1 Introduction
The use of an external power source (Chapter 7) to facilitate the cathodic oxygen
reduction in the RBEC has forfeited the possibility of generating net electrical energy as a by-
product. To increase the commercial promise of the proposed RBEC configuration, the process
should aim at converting the organic waste into a valuable by-product such as hydrogen.
Generating hydrogen from the cathode of the RBEC can be easily achieved with its enclosed
headspace design. Hydrogen should by theory be produced when a suitable cathodic potential is
provided by an external power source under anoxic condition. However, the operation of
membrane-less MEC for hydrogen production is always associated with the generic problems of
low hydrogen purity and the risk of hydrogen loss through hydrogenotrophic methanogenesis
(Call and Logan 2008; Clauwaert et al. 2008b; Logan et al. 2008). Different mitigation methods
have been proposed to optimize hydrogen yield by suppressing methane formation, but none of
these methods could eliminate methane formation in MEC.
In fact, methane formation in membrane-less MEC configurations is a very robust
process. Completely eliminating methane formation is therefore a very challenging task. On the
other hand, eliminating hydrogen formation to facilitate a high rate conversion of organics into
methane seems to be an easier option. From an energy recovery standpoint, methane could be a
more suitable energy carrier compare to hydrogen as it has a higher volumetric energy density.
Furthermore, compared to methane hydrogen has a much lower density and boiling point, only
0.09 kg Nm-3
and -252.9 °C vs. 0.717 kg Nm-3
and -182.5 °C for methane. These physical
properties make hydrogen gas unsuitable for transport and storage compared to methane, which
already has existing infrastructure to deal with its mass production. Hence, methanogenesis
could be an alternative option for practical operation of MEC.
In contrast to the previous chapter where the overpotential of oxygen at the cathode
severely limited the reaction, air was not supplied to the gas phase of the reactor such that the
cathodic halves of the disks of the RBEC reduce protons to hydrogen. However, it has been
shown in the literature that biofilms in the presence of hydrogen producing electrodes enrich for
the development of methanogenic micro-organisms (Ajayi et al. 2010; Clauwaert and Verstraete
2009; Jeremiasse et al. 2010). As the inhibition of methanogens is practically cumbersome and
the volumetric energy content of methane is higher than of hydrogen, this chapter aims at
producing methane as a byproduct from the treatment of dilute synthetic wastewater using the
RBEC.
Chapter 8: RBEC enables Electrochemically Driven Methanogenesis
- 144 -
8.2 Experimental Section
8.2.1 Bacterial Seeding Inoculum and Synthetic wastewater
A return activated sludge collected from a local domestic wastewater treatment plant
(sequencing batch reactors activated sludge process) was used as the initial inoculum. It was
stored at 4°C prior use. A synthetic wastewater was used throughout the study. Unless otherwise
stated, it consisted of (mg L-1
): NH4Cl 125, NaHCO3 125, MgSO4·7H2O 51, CaCl2·2H2O 300,
FeSO4·7H2O 6.25, and 1.25 mL L-1
of trace element solution, which contained (g L-1
): ethylene-
diamine tetra-acetic acid (EDTA) 15, ZnSO4·7H2O 0.43, CoCl2·6H2O 0.24, MnCl2·4H2O 0.99,
CuSO4·5H2O 0.25, NaMoO4·2H2O 0.22, NiCl2·6H2O 0.19, NaSeO4·10H2O 0.21, H3BO4 0.014,
and NaWO4·2H2O 0.050. Sodium acetate (5 to 10 mM) was added as the electron donor. Yeast
extract (1 g L-1
final concentration) was added as bacterial growth supplement. Unless otherwise
stated, the initial pH of the synthetic wastewater was adjusted to 6.9-7.2 using either 1M HCl or
4M NaOH.
8.2.2 Anoxic Operation of the Reactor Headspace of the Electrochemically-
Assisted RBEC
The enclosed headspace design of the RBEC reactor offers opportunity to recycle,
recover or control of gaseous byproducts. It is possible that by maintaining the headspace under
anoxic condition and applying a suitable external electrical power in a way similar to MEC
processes, the cathode of the RBEC reactor can generate hydrogen gas, an added-value
byproduct that is of commercial interest. All settings are similar to the system described in
Section 7.2.4.2 in Chapter 7, except that the headspace was completely isolated from the
atmospheric air throughout the experiment (Figure 8.1) (Photo 8.1). To obtain anoxic condition
in the headspace, the headspace was purged with pure nitrogen gas (1 L min-1
) for at least 10
min. In general, the influent has a composition as described in Section 7.2.3. It was continuously
fed into the reactor at a flow rate ranged from 1 to 3 ml min-1
. These flow rates corresponded to
hydraulic retention time (HRT) ranged from 9.7 to 19.4 h. COD removal rate was calculated
according to: (CODIn
-CODOut
)/HRT, where CODIn
and CODOut
are COD (mg L-1
) of the
influent and effluent, respectively. COD was measured according to a standard method (Closed
Reflux Colorimetric Method 1992)
Chapter 8: RBEC enables Electrochemically Driven Methanogenesis
- 145 -
Figure 8.1 A schematic diagram of the anoxic RBEC system for hydrogen or methane gas
production operated at a chemostat mode. (1) computer monitoring and control using
LabVIEW™; (2) gas counter ; (3) a computer controllable relay switch interfacing the
half-discs and the potentiostat; (4) peristaltic pump; (5) influent feed reservoir, at ~4°C
ice bath; (6) heat exchanger to warm up the cold feed; (7) the RBEC reactor unit; (8) gas
purging or sampling port; (9) reference electrode; (10) rotatable shaft and electrode disc
assembly; (11) computer controllable stepper motor; (12) effluent holding tank, at ~4°C
ice bath. Note: figure not drawn to scale.
Photo 8.1 Anoxic RBEC reactor for hydrogen or methane gas production operated at a
continuous mode. Heating tubing coil surrounding the reactor vessel was used to
maintain a constant temperature of 35°C.
Chapter 8: RBEC enables Electrochemically Driven Methanogenesis
- 146 -
8.2.3 Gas Production and Measurements
Biogas production from the RBEC was continuously recorded with a gas counter
connected to the headspace of the reactor. Prior to each experiment, the gas counter was
calibrated by injecting a known volume of N2 into the reactor to determine the gas volume per
count. This allows the calculation of the biogas production rate (BPR) (L biogas L-1
reactor day-
1). CH4, H2 and CO2 volumetric fractions (v/v, %) in the headspace were determined using a
Varian Star 3400 Gas Chromatograph (GC) coupled with a thermal conductivity detector. The
carrier gas was high purity nitrogen at a flow rate of 30 ml min-1
. Separation of different gas
species in the sample was performed on a Pora-PakQ packed column (2m x 5mm (internal
diameter)) maintained at 40oC. Detector and inlet temperatures were maintained at 40 and
120°C, respectively. Gas sample was taken from the reactor headspace via a gas-tight rubber
bung using a 10-mL plastic disposable syringe. About 50-100 µl of the gas sample inside the
plastic syringe were manually injected into the GC using a 100 µl gas-tight syringe (Hamilton).
Standard curves of each measured gas were established by injecting a known volume of high
purity standard gas into the GC.
The specific methane or hydrogen production rate (MPR or HPR) (L CH4 or H2 L-1
reactor day-1
) were calculated from the BPR and the volumetric fractions (v/v, %) of each gas in
the headspace (CH4headspace
or H2headspace
) according to the following equations:
MPR =CH4
headspace
100% × BPR (e.q. 8-2)
HPR =H2
headspace
100% × BPR (e.q. 8-3)
8.2.4 Examination of Electrode Biofilm using Scanning Electron Microscopy
The morphology of the biofilm growth at the electrode was examined by using scanning
electron microscopy (SEM) (Philips XL 20 Scanning Electron Microscope). At appropriate time
points as specified in the text, a small piece of carbon sheet was cut with a sterile stainless steel
scissor. The electrode samples were fixed in 3 % glutaraldehyde in 0.025 M phosphate buffer
Chapter 8: RBEC enables Electrochemically Driven Methanogenesis
- 147 -
(pH 7.0), overnight at 4 oC. The samples were then rinsed three times (5 min each time) with
0.025 M phosphate buffer (pH 7.0) before they were subjected to dehydration procedure through
a series of ethanol solution (each solution was changed two times and each change was 15 min):
30, 50, 70, 90 and 100%. The 100% ethanol solution was replaced with amyl acetate (two
changes, 15 min each. This step was performed in fume hood). The samples were critical point
dried before they were mounted to SEM specimen holder with contact adhesive. The samples
were then sputter coated with gold before being subjected to SEM observation. Abiotic plain
carbon sheet were treated with the same procedures and served as the Control.
Chapter 8: RBEC enables Electrochemically Driven Methanogenesis
- 148 -
8.3 Results and Discussion
The described RBEC was constructed, inoculated with activated sludge and operated
with an air-cathode for about three months. To establish a good anodic biofilm, a potentiostat
was connected to maintain the anodic potential at a suitable level (-300 to 0 mV) for the
anodophilic bacteria. Periodic half-turn rotation of the shaft enabled each electrode half disc to
intermittently serve as the submerged anode and the air-exposed cathode. This established a
biofilm capable of catalyzing the anodic as well as cathodic reaction (Chapter 7).
To test whether the developed reactor could also produce hydrogen as useful energy, the
oxygen in the headspace was replaced by nitrogen and its effect was recorded.
8.3.1 Anoxic Cathode enables Hydrogen Gas Formation
Switching the system from an oxygen reducing cathode (air in cathodic chamber) to an
anoxic cathode did not affect the electron flow (Figure 8.2). The potentiostat maintained a
constant anodic potential of -200 mV and hence the electron delivery by the bacteria to the
anode also stayed constant. However, switching from aerobic to anoxic headspace immediately
decreased the cathodic potential from about -1100 to -1300 mV (Figure 8.2B). Resuming the
headspace composition to air immediately increased the cathodic potential to the original level.
Under anoxic condition, the electrons at the cathode can only be accepted by electron
acceptors other than oxygen. In the literature, protons are commonly observed as the terminal
electron acceptor under such anoxic condition, leading to the cathodic hydrogen formation
according to equation 8-4:
2H+ + 2e
- → H2 E
o‘ = -612 mV (pH 7, 25°C) eq. 8-4
O2 + 4H+ + 4e
- → 2H2O E
o‘ = +618 mV (pH 7, 25°C) eq. 8-5
The reduction potential of this reaction is -612 mV while the reduction potential of the
oxygen reduction reaction is +618 mV (equation 8-5) (assume pH 7, 25°C). The results suggest
that the hydrogen producing reaction in the RBEC has about a 2.5 times lower overpotential
compared to the cathodic oxygen reduction.
Chapter 8: RBEC enables Electrochemically Driven Methanogenesis
- 149 -
Figure 8.2 Effect of different compositions in the reactor headspace on (A) current and the (B)
cathode potential (counter electrode potential). Submerged anode discs were poised at -
200 mV vs. Ag/AgCl. No flipping of electrode disc throughout the experiment. Solid
arrows indicate the onsets of purging with the test gas at 1 L min-1
. N2= 100% nitrogen;
CO2:N2= 20:80% carbon dioxide and nitrogen mix.
Cathodic hydrogen formation instead of oxygen reduction requires a cathodic potential
that is 200 mV lower. This would incur an increased power input in the process. However, with
hydrogen an additional fuel is produced, that may potentially offsets some of the electrical
energy input.
After switching the headspace from aerobic to anaerobic, hydrogen generation in the
RBEC was monitored at different applied anodic potentials, 0, -200 and -450 mV (Figure 8.3).
Over an 80 h period, each electrode half disc served alternately as an anode or a cathode upon an
hourly half rotation of the electrode disc assembly. The result shows that hydrogen production
was directly proportional to the current density (Figure 8.4). This observation is in line with
others using MECs for hydrogen production (Logan et al. 2008).
0
20
40
60
80
Cu
rre
nt (m
A)
-1500
-1300
-1100
-900
-700
-500
-300
0 10 20 30 40 50 60
Time (min)
Po
ten
tia
l (m
V v
s. A
g/A
gC
l)
.
Submerged Anode
Air-Exposed Cathode
Air Air AirCO2:N2 N2N2
CO2:N2
Chapter 8: RBEC enables Electrochemically Driven Methanogenesis
- 150 -
Figure 8.3 Effect of anode potential on (A) hydrogen production; current and (B) electrode
potentials in the RBEC operated under anoxic cathodic condition. The system was
switched from an air-cathode to anoxic cathode for about two days prior to the
experiment. Discs were flipped once per hour. Submerged disc = 1 and 0 indicate half
disc 1 and 2, respectively.
Figure 8.4 Hydrogen production rate as a function of current density in the RBEC operated
under anoxic cathodic condition. Data are calculated from Figure 8.3.
y = 0.04x - 0.01
R2 = 0.99
0.00
0.02
0.04
0.06
0.0 0.5 1.0 1.5 2.0
Current Density (A m-2
)
Hyd
rog
en
Pro
du
ctio
n R
ate
.
(m3 m
-3 d
-1)
(A)
(B)
Chapter 8: RBEC enables Electrochemically Driven Methanogenesis
- 151 -
Table 8.1 Effect of anodic potential set point on the reaction rate and net energy input of the
RBEC reactor.
Poised Anodic Potential
I COD Removal Rate
H2 Production Rate
Electricity Input per COD Removed
Energy Recovered as H2 per COD Removed
Net Electricity Input per COD Removed
a
mV mA kg ·m-3
·d-1
m3·m
-3·d
-1 kWh·kg
-1 kWh kg
-1 kWh·kg
-1
0 90 0.35 0.06 2.66 0.45 2.21
-200 72 0.26 0.05 2.12 0.50 1.62
-450 41 0.17 0.003 0.75 0.05 0.70
Data were obtained during the first 4 days after switching from aerobic to anaerobic cathodic headspace.
Energy content of hydrogen is based on its heat of combustion value, -285.8 kJ·mol-1
aOnly the energy input from the potentiostat is considered here.
On the one hand the provision of external electrical energy to ―push the reaction‖ would
be an operating cost factor, on the other hand it speeds up the reaction and hence would enable
to use smaller vessel sizes (lower capital costs). To visualize the trade-off between energy input
and reaction rate the reactor was operated at three different anodic potentials (Table 8.1). Results
show that a significant energy cost would be incurred when operating the reactor and the COD
removal rates are lower than that obtained from activated sludge plants, 0.5-2 kg COD·m-3
·d-1
(Logan 2008).
8.3.2 Methane instead of Hydrogen was the Predominant By-product after
Prolonged Anoxic Operation
After operating the RBEC reactor for over two weeks under fully anoxic condition, up to
48% of methane was detected at the headspace instead of hydrogen (0.1%). SEM photos of the
biofilm electrode collected from the methane-producing RBEC illustrated that a well developed
biofilm was developed at the carbon electrode compared to the abiotic control (Figure 8.5).
Methane generation was also observed by other authors and could not be avoided in membrane-
less system (Cheng et al. 2009b; Clauwaert and Verstraete 2008; Hu et al. 2008). Rather than
attempting to suppress methanogenesis it was aimed at using and maximizing this reaction and
produce methane to offset the energy input.
As observed for hydrogen the rate of methane formation of the RBEC was directly
proportional to both the current and the electrical energy input (Figure 8.6A and B). With an
average current density of 4.1 A m-2
(390 mA) (the anodic and cathodic potentials were -200 and
Chapter 8: RBEC enables Electrochemically Driven Methanogenesis
- 152 -
-1600 mV, respectively), the maximal COD removal rate and methane production rate were 1.38
kg COD m-3
day-1
and 0.53 L L-1
day-1
, respectively (Figure 8.6A). However, when the current
was reduced to less than 0.16 A m-2
(15 mA) (the anodic and cathodic potentials were -560 and -
850 mV, respectively), the COD removal rate and methane production rate were remarkably
reduced to 0.2 kg COD m-3
day-1
and 0.04 L L-1
day-1
, respectively (Figure 8.6A).
Figure 8.5 Scanning electron micrographs of (A to E) a carbon fiber sheet attached with the
biofilm obtained from the methane-producing RBEC under alternate disc flipping
operation (anoxic headspace) at different magnification; and (F) a plain carbon
fiber sheet. The samples were collected after about two weeks the reactor was
switched from aerobic cathode into fully anoxic operation. Dotted circles in A to D
indicate the same location of the same sample.
A B
C D
E F
5 µm
50 µm 5 µm
500 µm 200 µm
50 µm
Chapter 8: RBEC enables Electrochemically Driven Methanogenesis
- 153 -
Figure 8.6 Methane production rate and COD removal rate as a function of (A) current and (B)
electricity input in the RBEC operated under anoxic condition. Temperature was
controlled at 35°C.
When voltage was not applied and the external circuit was shorted (i.e. close to zero
resistance between the upper and lower half discs), the methane production was only 0.03 L L-1
day-1
(data not shown). This low, non-current driven methane production could be attributed to
the activity of acetoclastic methanogens in the system and/or that syntrophic acetate oxidation
coupled with hydrogenotrophic methanogenesis (Clauwaert and Verstraete 2008). Overall, the
results demonstrate that the current-dependent methane formation was not triggered by the
anode biofilm, but was likely caused by the electron donating cathode.
At cathodic potentials of -1200 and -1600 mV, 82 and 88% of the electron flow in the
RBEC ended up as methane gas (Figure 8.7). While at a cathodic potential of -850 mV (<15
mA), the current could not explain the methane production. This can be due to the fact that the
electrons have by-passed the current generation via alternative methane production pathway(s).
R2
= 0.99R
2= 0.99
0
0.2
0.4
0.60 100 300 400Current (mA)
0
0.5
1.0
1.5Methane Production Rate
COD Removal Rate
(B)
(A)
R2
= 0.92
R2
= 0.98
0
0.2
0.4
2 4 6Electrical Energy Input (kWh kg COD-1)
Me
tha
ne P
rod
uctio
n R
ate
(L L
-1d
ay
-1)
0
0.5
1.0
CO
D R
em
ova
l Ra
te(k
g C
OD
m-3
da
y-1)
Chapter 8: RBEC enables Electrochemically Driven Methanogenesis
- 154 -
Figure 8.7 Electron balance (as electron flow rates) in the methane-producing RBEC with a
cathode poised at -850, -1200 or -1600 mV.
8.3.3 Is Hydrogen a Key End-product of the Cathodic Reduction Reaction in the
Methane-Producing RBEC?
From a thermodynamic perspective, if hydrogen generation (equation 8-4) is the only
cathodic reaction, then for every 10-fold increase in hydrogen partial pressures inside the
headspace the electrode potential is expected to decrease by about 30 mV (Figure 8.8).
Increasing the end product concentration (here H2) would lead to a linear increase of the Gibbs
free energy change in the reaction, rendering the reaction energetically less favorable. On the
contrary, it is expected that by reducing the hydrogen partial pressure in the cathodic headspace,
the cathodic hydrogen production reaction would become more energetically favorable.
However, such phenomenon was not observed in the RBEC reactor. The result shows
that even a 200-fold increase of hydrogen partial pressure in the headspace did not lead to either
a more negative electrode potential or a decrease in current (Figure 8.9). Hence, hydrogen is
unlikely the key cathodic end product that linked the electron transfer from the cathode to
methane production. In the literature, an evidence of methanogenic microorganisms directly
acquire electrons from a cathode to produce methane is recently published (Cheng et al. 2009b).
Further study is required to elucidate the exact mechanism of the cathodic methanogenesis in the
RBEC.
0
50
100
150
200
Electrical
Current
COD
Removal
Hydrogen
Production
Methane
Production
Ele
ctr
on F
low
Ra
te (
mm
ole
L-1
d-1
) -850 mV -1200 mV -1600 mV
0
50
100
150
200
Electrical
Current
COD
Removal
Hydrogen
Production
Methane
Production
Ele
ctr
on F
low
Ra
te (
mm
ole
L-1
d-1
) -850 mV -1200 mV -1600 mV
Chapter 8: RBEC enables Electrochemically Driven Methanogenesis
- 155 -
Figure 8.8 Gibbs free energy change (ΔG°‘) and electrode potential (ΔE°‘) of hydrogen
production reactions (2H+ + 2e
- → H2) at different hydrogen partial pressures.
Values of ΔG°‘ were calculated from (Thauer et al. 1977); ΔE°‘ were calculated
from ΔG°‘ according to ΔG°‘= nFΔE, where n is mole of electron transfer per H2
formed, F is the Faraday‘s constant, 96485 C mol-1
.
Figure 8.9 Current production at different cathodic potentials of the RBEC with either N2
(open symbols) or H2 (solid symbols) in the headspace of the cathode. Note: Prior
to the experiments, the headspace was flushed with pure N2 (1L min-1
) for 5 min.
-100
-80
-60
-40
-20
0
20
40
60
80
100
-30 -28 -26 -24 -22 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0
Hydrogen Partial Pressure (Log10 atm)
∆G 0
' (kJ m
ol-1
)
-700
-500
-300
-100
100
300
∆E 0 ' (m
V v
s. A
g/A
gC
l)
Gibbs free energy for H2 production
Electrode Potential for H2 production
0
20
40
60
80
100
-1350-1150-950-750
Cathodic Potential (mV vs. Ag/AgCl)
Cu
rre
nt (m
A)
>99% H2
>99% N2 (~0.5% H2)
Chapter 8: RBEC enables Electrochemically Driven Methanogenesis
- 156 -
8.3.4 Energy Balance
Two approaches were used to calculate the energy recovery from the methanogenic
RBEC process. They are both based on the energy content of the methane recovered, compared
to (1) the electrical energy input plus the energy content in the substrate (i.e. as COD) removed;
and (2) the electrical energy input only (Logan et al. 2008). With the first approach, the energy
recoveries obtained from the three applied cathodic potential settings of -850, -1200 and -1600
mV are 43, 44 and 46%, respectively. The energy content of the organic pollutants in a
wastewater can be considered as a low quality, ―useless‖ energy. Therefore, it is neglected here
in our energy calculation as in the case for conventional anaerobic digestion which usually yield
net energy gain. With this second approach, the energy recoveries obtained from the three
applied cathodic potential settings of -850, -1200 and -1600 mV are 686, 93 and 75%,
respectively.
Figure 8.10 Energy balance (energy recovered as CH4 minus electrical energy input) and COD
removal rate (COD RR) as a function of applied voltage in the methane-producing
RBEC. *Value was re-calculated from this reference as: applied voltage= -825 ±
12 mV; energy balance= -0.14 ± 0.35 kWh kg COD-1
.
The estimated energy balance (energy recovered as CH4 minus electrical energy input) of
the methane-producing RBEC is directly correlated with the applied voltage (Figure 8.10). The
higher the applied voltage, the lower the energy recovery. This result also agrees with Clauwaert
and Verstraete (2008) that a methanogenic bioelectrochemical system may give a negative
y = 0.003x + 2.219
R2 = 0.977
y = -0.001x - 0.097
R2 = 0.993
-2.0
-1.0
0.0
1.0
2.0
-1500-1000-5000
Applied Voltage (mV)
Ou
tpu
t CH
4 -
In
pu
t Ele
c (k
Wh
kg
CO
D-1
)
/ C
OD
RR
(kg
CO
D m
-3 d
ay-1
)
Methane-producing RBEC
Clauwaert and Verstraete (2008)*
COD Removal Rate
Chapter 8: RBEC enables Electrochemically Driven Methanogenesis
- 157 -
energy yield (-0.14 ± 0.35 kWh kg COD-1
). To obtain a positive energy yield one would come at
a cost of a lowered COD removal rate (here 0.2 kg COD m-3
day-1
) (the triangle symbols in
Figure 8.10). Certainly, from a wastewater treatment standpoint this is an undesirable outcome.
Decision needs to be justified on a basis of what the process aims to accomplish.
8.4 Implication of Findings
This study demonstrated that methane rather than hydrogen was the predominant
electron sink of the RBEC operated under anoxic condition. The results are in line with Cheng et
al. (2009) that the methane formation rate was proportional to the cathodic potential and the
applied voltage in a BES. To our current understanding, methane production under reactor
conditions described in the present study can only be of biological but not of abiotic nature. This
implies that both the anodic and cathodic reactions occur in the RBEC were biologically
catalyzed at the electrodes without using chemical catalysts such as platinum. However, whether
the methane producing bacteria can accept electrons directly from the cathode for a one-step
methane formation still remains unclear. Further studies are needed to verify the mechanism of
the bioelectrochemical methane generation and to optimize the RBEC for the proposed
application.
- 158 -
9 Conclusions and Outlook
Chapter Summary
Overall, this thesis has explored the potential of bioelectrochemical systems for energy
recovery from wastewater. It has extended our fundamental understanding on how
electrochemically active microorganisms behave and responsed to the unique environment in
BES. From a practical perspective, the new RBEC configuration may widen the functionality or
suitability of BES for a large-scale wastewater treatment application. However, it remains a
challenge to justify the applicability of BES as a direct one-step wastewater-to-electricity
technology.
This final chapter discusses the potential of bioelectrochemical systems for energy
recovery from wastewater based on the knowledge gained in this thesis. The insights obtained in
this thesis are summarized and discussed. The limitations of the thesis and recommendation for
future research are also addressed.
Chapter 9: Conclusions and Outlook
- 159 -
9.1 Conceptual Progression of the Thesis
9.1.1 A highly Anodophilic Biofilm can be Easily Established from Activated Sludge
This thesis began with evaluating the electricity-producing capacity of an established
electroactive microbial consortium in a classical two-chamber laboratory bioelectrochemical cell
(Chapter 2). An easily obtainable, non-specific mixed culture inoculum (activated sludge from a
wastewater treatment plant) could easily evolve into an anodophilic biofilm on a non-chemically
catalyzed granular graphite anode (bioanode). Similar to most other studies, artificial electron
shuttles (mediators) were not required in the microbial acclimation process. This is a crucial
microbial property because both maintaining sterility and adding artificial electron mediators
(normally they are colored, toxic chemicals) are undesirable requirements for practical BES
operation dealing with wastewater.
Under a well-maintained anodic condition (with the aid of a computer-feedback control),
the established bioanode could effectively generate electricity while removing the organic
pollutant from a wastewater. The established favorable condition included: high buffer strength
and neutral pH in the anolyte (wastewater); efficient mass transfer (provided by a high
recirculation rate of the anolyte to minimize concentration gradient) and an effective ―electron
accepting‖ anode. The low biomass yield (high specific activity) of the bioanode is another merit
of using bioanode for wastewater treatment as it may represent a cost saving in sludge post-
treatment. Nonetheless, the key obstacle of up-scaling (or commercializing) MFC for practical
wastewater treatment application does not lie in the bioanode process.
9.1.1.1 MFC Power Output is undermined by Limitations other than the Metabolic
Capacity of the Bioanode
If we consider only the bioanode capacity and assuming that all other processes in the
MFC (e.g. ionic or ohmic resistance, electrolyte pH and cathodic reduction) are non-limiting, the
potentially retrievable electrical power output is actually more than enough to justify a large-
scale MFC process. According to Rabaey and Verstraete (2005), practical MFC processes aimed
at energy recovery should reach a power output of ≥ 1 kW·m-3
(as compared to conventional
anaerobic digestion technology). This target level has been achieved in Chapter 2 using the
computer-controlled, ferricyanide-cathode MFC in which the capacity of the anodophilic biofilm
would be fully exploited.
Chapter 9: Conclusions and Outlook
- 160 -
Let us consider only the capacity of the established bioanode to anticipate power output
and COD removal in an up-scaled (25 m3) MFC reactor (Figure 9.1). We assume that
technologies could be developed that overcome the generic limitations such as the poor cathodic
oxygen reduction and other electrochemical losses of the process (Box 9.1).
Figure 9.1 Stacking up individual microbial fuel cells in-series to amplify voltage and power
output in a hypothetical up-scaled system. The diagram is not drawn to scale.
Although the estimated COD removal rate is significantly lower compared to the highest
COD removal rate (45 kg COD m-3
d-1
) of the best anaerobic digester available for today
(Expand Granular Sludge Bed) (van Lier et al. 2008), it is well beyond (3-fold) the benchmark
level of 10 kg COD m-3
d-1
for practical MFC up-scaling (Clauwaert et al. 2008a; Pham et al.
2009; Rabaey and Verstraete 2005). Nonetheless, such a promising performance has never been
achieved under realistic condition as very often the capacity of a bioanode is undermined by
other generic limitations such as the poor cathodic oxygen reactivity and the proton
concentration polarization between the two opposing electrodes (pH gradient or pH split).
Addressing these limitations often requires impractical methods. As in Chapter 2, high
performance was obtained by using a ferricyanide-cathode (for effective cathodic reaction), a
high strength phosphate buffer and acid/base dosing (pH control). Although the concept of
polarity inversion invented in Chapter 6 may potentially overcome these generic limitations, it
CO
D
CO
D
CO
D
CO
D
CO
D
CO
D
CO
D
CO
D
CO
D
CO
D
0.1 m
5 m
5 m
2.5 m3
A
n
o
d
i
c
C
h
a
m
b
e
r
Anode
Cathode
Ion exchange membrane
Power Output = 50 kW
Chapter 9: Conclusions and Outlook
- 161 -
remains as a major research challenge to develop a system that can fully exploit the bacterial
capacity to achieve the benchmark target of ≥ 1 kW·m-3
. Perhaps, as already suggested by other
colleagues, the use of BES to recover or even generate other valuable by-products (hydrogen,
methane, 1,3 propanediol, ethanol or butanol…etc), rather than the direct recovery of electricity
from wastewater may have a better chance of commercial success in the technology market
(Rozendal et al. 2008a; Schroder 2008).
Box 9.1 Estimation of the power output and COD removal rate of an up-scaled MFC
wastewater treatment staked module by considering only the bioanode capacity.
In this estimation, we assume that…
10 MFC units are connected in-series as a stack
The bioanode at each unit could sustain an average volumetric current density of
4000A m-3, at an operating cell voltage of 500 mV (anodic potential -400 mV;
cathodic potential +100 mV) (Chapter 2)
100% coulombic efficiency
All other steps in the MFC are not limiting
Power Output:
Power = Voltage × Current
Voltage: Ten MFC units connected in-series gives: 500 × 10 mV = 5 V
Each MFC unit delivers a current of 10000 A (4000 A m-3 × 2.5 m3)
Hence, Power (W) = 5 V × 10000 A = 50 kW
Volumetric power density = 50 kW/ (2.5 × 10) m3
= 2 kW·m-3
COD Removal Rate:
The COD removal rate (here as acetate) can be estimated from the current:
Current delivered by the MFC stack = 4000A m-3 × 25 m3 = 100000A = 360000000 C h-1
= 3731 mol e h-1 (’ 96485; Faraday‘s constant)
Acetate removal rate = 466.4 mol Ac h-1 (÷ 8; 8 e-/ mole Ac)
= 27517 g Ac h-1 (× 59; 59g/ mole Ac)
COD removal rate = 29361 g COD h-1 (× 1.067; 1 g Ac = 1.067 g COD)
= 29.4 kg COD h-1 (÷ 1000; 1000 g/kg)
= 704.7 kg COD d-1 (× 24; 24 h/day)
= 28.2 kg COD m-3
d-1 (÷ 25; 25 m3/stacked MFC)
9.1.1.2 Wastewater Treating-BES requires Innovative R&D Approaches
Looking back to the past decade, the development of MFC technology was very much
inspired ―or influenced‖ by the conventional chemical fuel cell (e.g. PEM H2-FC) domain (e.g.
reactor configurations, sophisticated ion exchange membrane, membrane electrode assembly,
Chapter 9: Conclusions and Outlook
- 162 -
chemical catalysts, etc). To practically implement MFC for wastewater treatment application, we
may need to exploit completely new R&D directions and approaches.
It is worth noting that no matter to what extent we could overcome the limitations in a
MFC, MFC-based energy powering systems (e.g. automobile‘s engines) don‘t compare
favourably with conventional chemical fuel cells (e.g. PEM hydrogen fuel cell). As shown in
Figure 9.2, the current density (volumetric) of a typical MFC is at least four orders of magnitude
lower than that of a typical PEM H2 fuel cell.
Figure 9.2 The power output of conventional hydrogen fuel cell (PEMFC) far exceeds the
power output of microbial fuel cells. (PEMFC data is obtained from Mench et al. (2001);
MFC-Ferricyanide data is from Chapter 2; MFC-Dissolved O2 data is from Clauwaert et
al. (2008a)).
There are several properties of (domestic) wastewaters that may hinder their use as a
feedstock for electricity production in BES:
Low energy content. A large volumetric portion of wastewater consists of purely water
instead of the ―fuel‖ (i.e. organics) (Verstraete et al. 2009): Is that possible to run a
water-less BES fueled with 99% (v/v) organics?
Wastewater is heavy (> 1 kg/L). Distributing or re-circulating wastewater inside the BES
costs energy (mechanical). Taller BES systems may require more of this energy. Yet,
land area availability becomes a concern for a shorter/ flatter system (reactor footprint).
The design of RBEC configuration (Chapter 7) may help minimize the recirculation
requirement. In contrast, H2-driven chemical fuel cells do not face this problem as the
1300
2 0.1
0
200
400
600
800
1000
1200
1400
Po
we
r D
en
sity (
kW
m-3
)
PEMFC MFC-
Ferricyanide
MFC-
Dissolved O2
Not sustainable and practical; For research only
Maybe Sustainable
Chapter 9: Conclusions and Outlook
- 163 -
fuel is in gaseous form, which is light and also has higher energy content (e.g.
compressed H2).
Wastewaters are normally poorly conductive with only low pH buffering capacity. These
explain why by reducing the electrode spacing (i.e. distance between anode and cathode)
or by adding extra chemical buffer into the electrolyte (wastewater), power output of the
BES is enhanced.
9.1.2 Anodophilic Biofilm Affinity for the Anodic Potential
Bioanode is considered as a common component for both MFC and MFC. Hence,
understanding the bacterial behavior of bioanode is of fundamental importance to BES
development. A generic relationship between microbial activity and anodic potential has been
established in Chapter 3. Two parameters, the half-saturation anodic potential (kAP) and the
critical anodic potential (APcrit.) are defined to describe the affinity of an anodophilic biofilm for
the anodic potential. They are expected to aid the quantitative evaluation or modeling of MFC
anodic processes. A similar biofilm-anode affinity concept have been recently proven by others
using both experimental and modeling approaches (Lee et al. 2009; Marcus et al. 2007; Torres et
al. 2007).
From a practical perspective, knowing the affinity of a specific electroactive culture may
be useful for selecting a specific anodic microbial community from a mixed consortium —
―Electrowinning of bacteria” concept (similar to ―Biofilm-plating‖ as suggested by Pham et al.
(2009)). For example, one could selectivity harvest a bacterial culture with a specifically
modified electrode surface by controlling the anode at a specific potential to ―pick up‖ the target
candidate(s) from a microbial consortia. This is a situation similar to the use of a huge
electromagnet (with controllable magnetic strength) to separate magnetized iron materials from
a heap of rubbish. Depending on the purpose of the ―electrowinning‖, the electrode could be first
coated with a unique mediator, enzyme and drug or even preloaded with another microbial
species or consortia.
9.1.3 The Use of Computer Feedback for Dynamic BES Process Control
Computer-based bioprocess control is particularly useful when up-scaling BES processes
for applications such as wastewater treatment. Similar concepts have been widely adopted in the
wastewater treatment industry (e.g. controlled aeration in a sequencing batch reactor (SBR)
based on dissolved oxygen concentration set points). However, in the literature, the use of
computer programs to control BES processes in a dynamic interactive manner is still uncommon.
Chapter 9: Conclusions and Outlook
- 164 -
This thesis has demonstrated the merit of using computer feedback for advanced
operation of a BES process. For example, feedback controlling the external resistance could
dynamically control the electrode potential and power output of a MFC, or even running a cyclic
voltammetry analysis (Chapter 4). Other examples are the alternating supply of substrate and
oxygen to a single biofilm for the polarity inversion study in Chapter 6 and the rotation of
electrode discs of the RBEC in Chapter 7 and 8.
Obviously, the application of computer feedback control deserves further research
advancement. Below are some potential merits it may offer to a large-scale BES wastewater
treatment process:
Process flexibility Different operational regimes can be tailored for specific
needs.
High precision Computer program can synchronize even highly
sophisticated operational demand.
Low maintenance, low
labor cost
BES is predominately controlled by the program; requires
only little attention from the operator.
Easy to gain knowledge on
system performance
Computer can record input (cause) and output (effect)
parameters (involved in the feedback control loop) over time.
Experienced operators can easily establish understanding on
the system for on-going process improvement, or to develop
database of the process.
9.1.4 The Rotatable Bioelectrochemical Contactor: A New Option for Large-Scale
Wastewater Treatment
The discovery that an electrochemically active biofilm could catalyze both anodic and
cathodic reactions in a BES leads to the development of a novel BES configuration, the RBEC.
Two operating modes, MFC and MEC have been tested. By allowing electron flow from the
water-immersed to the air-exposed portion of a biological contactor, organic removal rate can be
accelerated and electrical energy can be directly generated from the process as a MFC. The
described RBEC process has not been optimized yet. Similar to other fixed-film biological
Chapter 9: Conclusions and Outlook
- 165 -
wastewater treatment processes (e.g trickling filter), a significant improvement of both organic
removal and electricity generation is expected by increasing the surface area of the electrode per
reactor volume of the RBEC (e.g. by packing more electrode disc onto the central rotating shaft
without increasing the reactor volume) (Figure 9.3). This may support a higher microbial
activity in the reactor, which may increase the bioelectrochemical conversion of the organics
into electricity. However, this may impose a higher material cost of the process.
Figure 9.3 Hypothetical large-scale RBEC systems for (A) MFC and (B) MEC operations. More
electrode discs are packed inside the reactor to maximize specific surface area for
improved electrochemical reactions.
Cathode
Anode
CODInfluent
CODEffluent
Waste Sludge
Oxygen Oxygen Oxygen
Electricity Output
Recirculation (optional)
Cathode
Anode
CODInfluent
CODEffluent
Waste Sludge
Methane Rich BiogasElectricity Input
Recirculation (optional)
Enclosed Headspace
A.
B.
Chapter 9: Conclusions and Outlook
- 166 -
A possible drawback when scaling up the RBEC is that the ohmic resistance caused by
the ionic flux may become significant in a larger system. This is because the ions (either from
the submerged anode or the exposed cathode) may need to migrate for a longer distance. In
principle, a one-dimensional increase in the disc diameter would lead to a two-dimensional
increase in the disc surface area (Figure 9.4). Although the increased disc diameter would
increase the boundary for ionic flux across the upper and the lower discs, the exponential
increase in the electrode size (i.e. increased current) may potentially render the ionic flux to
become limiting. This concern should be a subject of a further study.
Figure 9.4 Ionic fluxes could be a potential up-scaling limitation of the RBEC: Increasing the
diameter of the RBEC electrode disc leads to exponential increase in electrode surface
area.
Ionic fluxes boundary between anodic and cathodic half discs.
Chapter 9: Conclusions and Outlook
- 167 -
9.1.4.1 RBEC as a MEC for enhanced cathodic oxygen reduction, hydrogen or
methane production
Enhanced Cathodic Oxygen Reduction. The organic removal could be further improved
by adding an extra power supply to the RBEC (up to 1.32 kg COD·m-3
·d-1
). Although net
electrical energy can no longer be obtained, the overall energy consumption is still lower
compared to the conventional activated sludge processes (Table 9.1).
Hydrogen or Methane Production? When the RBEC was operated under fully anaerobic
condition together with the external power supply, hydrogen or methane gas was produced.
Similar to other MECs, the gas production rate in the RBEC was dependent on the applied
power. It has to be noted that in order to produce hydrogen gas as the dominant gas product, a
suitable measure is required to suppress methane formation. The prolonged operation (weeks) of
the RBEC under anaerobic condition resulted in only methane generation. Further studies are
needed to verify the mechanism of the bioelectrochemical methane generation and to optimize
the RBEC for the proposed application.
Table 9.1. Treatment performance and energy comparison of different W.W.T. systems.
Process Status of the Technology Energy Recovery?
COD Removal Rate
Energy Consumption
kg COD·m
-3·d
-1 kWh·kg COD
-1
AS Developed in 1913 and is widely adopted nowadays.
0.5 to 2.0 a 0.7 to 2.0
a
MEC R&D stage (since 2005) as H2 ~ 6.5 a 0.5 to 2.4
a
RBEC (O2 reduction)
No prior art 1.32 0.47
RBEC (H2 production)
No prior art as H2
0.35 b 2.21
RBEC (CH4 production)
No prior art as CH4
0.2 to 1.38 -1.5 to 1.5
AS = activated sludge; MEC = microbial electrolysis cells; a values obtained from Logan et al. (2008)
Methane is a well-known, practical energy carrier in the wastewater industry as it has a
higher volumetric energy density than hydrogen. Using electricity to drive the RBEC process for
methane generation may represent a completely new way to conserve energy surplus from other
alternative renewable energy facilities (e.g. a solar panel or a wind-farm). For example, during a
sunny day where excess electricity is generated by a solar panel a RBEC can convert this energy
into methane while treating a wastewater. The RBEC can be switched back to a MFC-mode for
electricity generation when the energy supply is not available.
Chapter 9: Conclusions and Outlook
- 168 -
At this early stage, particular attention should be given to the optimization of cost
effective materials and the microbial functionality in this novel reactor configuration. Other
considerations such as cost and life-cycle assessments may be required at a later stage to justify
the applicability of the RBEC in a real world situation.
9.2 Final Remarks
Overall, the thesis has explored, from both a fundamental and practical perspectives, the
potential of bioelectrochemical systems for energy recovery from wastewater. Fundamentally,
this thesis has extended our understanding on how electrochemically active microorganisms
behave and response to the unique environment in BES. Especially, the discovery of the
bidirectional microbial electron transfer property may shed light not only on BES development,
but also on the context of fundamental microbiology. Practically, the new RBEC configuration
may widen the functionality or suitability of BES for a large-scale wastewater treatment
application. However, it remains a challenge to justify the applicability of BES as a direct one-
step wastewater-to-electricity technology. Although the noval operational regime of intermittent
prolarity inversion developed in this thesis could be beneficial to the well-known pH gradient
limitation, further process optimization of a practical BES process is still needed. Besides, the
cathodic reaction still remains as the key process bottleneck. Future research efforts should go
towards improving and elucidating in details the mechanisms of the microbe-cathode electron
transfers.
References
- 169 -
R References
Adams, M.W.W., Stiefel, E.I., 1998. Biological hydrogen production: not so elementary.
Science 282(5395), 1842-1843.
Aelterman, P., Freguia, S., Keller, J., Verstraete, W., Rabaey, K., 2008. The anode potential
regulates bacterial activity in microbial fuel cells. Appl Microbiol Biotechnol 78(3), 409-418.
Ajayi, F.F., Kim, K.-Y., Chae, K.-J., Choi, M.-J., 2010. Effect of hydrodymamic force and
prolonged oxygen exposure on the performance of anodic biofilm in microbial electrolysis cells.
International Journal of Hydrogen Energy 35(8), 3206-3213.
Angenent, L., Karim, K., Al-Dahhan, M., Domiguez-Espinosa, R., 2004. Production of
bioenergy and biochemicals from industrial and agricultural wastewater. Trends Biotechnol.
22(9), 477-485.
Bard, A.J., Faulkner, L.R., 2001. Electrochemical methods: fundamentals and applications, 2 ed.
John Wiley & Sons, Inc., New York, USA.
Bard, A.J., Parsons, R., Jordan, J., 1985. Standard potentials in aqueous solution. Marcel Dekker,
New York.
Benemann, J., 1996. Hydrogen biotechnology: Progress and prospects. Nat. Biotechnol. 14,
1101-1103.
Bennetto, H.P., Stirling, J.L., 1983. Anodic reactions in microbial fuel cells. Biotechnol. Bioeng.
25, 559-568.
Biffinger, J., Ribbens, M., Ringeisen, B., Pietron, J., Finkel, S., Nealson, K., 2008.
Characterization of electrochemically active bacteria utilizing a high-throughput voltage-based
screening assay. Biotechnol. Bioeng. DOI:10.1002/bit.22072.
Boeriu, C.G., van Dam, J.E.G., Sanders, J.P.M., 2005. Biomass valorisation for sustainable
development. In: Lens, P., Westermann, P., Haberbauer, M., Moreno, A. (Eds.), Biofuels for
fuel cells: renewable energy from biomass fermentation, pp. 17-34. IWA Publishing, London,
UK.
Bond, D.R., Lovley, D.R., 2003. Electricity production by Geobacter sulfurreducens attached to
electrodes. Appl. Environ. Microbiol. 69(3), 1548-1555.
Bond, D.R., Lovley, D.R., 2005. Evidence for involvement of an electron shuttle in electricity
generation by Geothrix fermentans. Appl. Environ. Microbiol. 71(4), 2186-2189.
References
- 170 -
Bossel, U., 2006. Does a hydrogen economy make sense? Proc. IEEE 94(10), 1826-1837.
Bryant, M.P., Wolin, E.A., Wolin, M.J., Wolfe, R.S., 1967. Methanobacillus omelianskii, a
symbiotic association of two speices of bacteria. Arch. Microbiol. 59, 20-31.
Call, D., Logan, B.E., 2008. Hydrogen production in a single chamber microbial electrolysis cell
lacking a membrane. Environ. Sci. Technol. 42(9), 3401-3406.
Chachuat, B., Roche, N., Latifi, M.A., 2005. Long-term optimal aeration strategies for small-
size alternating activated sludge treatment plants. Chem. Eng. Process. 44, 593-606.
Chamberlain, J.C., Foley, H.M., MacDonald, G.J., Ruderman, M.A., 1982. Carbon dioxide
review 1982. Oxford University Press, New York, USA.
Chambers, B., Upton, J., Greenhalgh, S.H., 1998. The development and use of hybrid aeration
processes. Wat. Environ. J. 12, 322-329.
Chaudhuri, S.K., Lovley, D.R., 2003. Electricity generation by direct oxidation of glucose in
mediatorless microbial fuel cells. Nat. Biotechnol. 21(10), 1229-1232.
Chen, E., 2003. Thermodynamics and electrochemical kinetics. In: Hoogers, G. (Ed.), Fuel cell
technology handbook. CRC Press LLC.
Cheng, K.Y., Cord-Ruwisch, R., Ho, G., 2007. Limitations of bio-hydrogen production by
anaerobic fermentation process: an overview. American Institute of Physics Conf. Proc. 941,
264-269.
Cheng, K.Y., Cord-Ruwisch, R., Ho, G., 2009a. A new approach for in situ cyclic voltammetry
of a microbial fuel cell biofilm without using a potentiostat. Bioelectrochemistry 74(2), 227-231.
Cheng, K.Y., Ho, G., Cord-Ruwisch, R., 2008. Affinity of microbial fuel cell biofilm for the
anodic potential. Environ. Sci. Technol. 42(10), 3828-3834.
Cheng, S., Xing, D., Call, D.F., Logan, B.E., 2009b. Direct biological conversion of electrical
current into methane by electromethanogenesis. Environ. Sci. Technol. 43, 3953-3958.
Choi, Y., Kim, N., Kim, S., Jung, S., 2003. Dynamic behaviors of redox mediators within the
hydrophobic layers as an important factor for effective microbial fuel cell operation. Bull.
Korean Chem. Soc. 24(4), 437-440.
Claassen, P.A.M., van Lier, J.B., Contreras, A.M.L., van Niel, E.W.J., Sijtsma, L., Stams,
A.J.M., de Vries, S.S., Weusthuis, R.A., 1999. Utilisation of biomass for the supply of energy
carriers. Appl. Microbiol. Biotechnol. 52, 741-755.
Clauwaert, P., Aelterman, P., Pham, H.T., De Schamphelaire, L., Carballa, M., Rabaey, K.,
Verstraete, W., 2008a. Minimizing losses in bio-electrochemical systems: the road to
applications. Appl. Microbiol. Biotechnol. 79, 901-913.
References
- 171 -
Clauwaert, P., Mulenga, S., Aelterman, P., Verstraete, W., 2009. Litre-scale microbial fuel cells
operated in a complete loop. Appl. Microbiol. Biotechnol. 83(2), 241-247.
Clauwaert, P., Rabaey, K., Aelterman, P., Schamphelaire, L.D., Pham, T.H., Boeckx, P., Boon,
N., Verstraete, W., 2007a. Biological denitrification in microbial fuel cells. Environ. Sci.
Technol. 41(9), 3354 -3360.
Clauwaert, P., Toledo, R., van der Ha, D., Crab, R., Verstraete, W., Hu, H., Udert, K.M., Rabaey,
K., 2008b. Combining biocatalyzed electrolysis with anaerobic digestion. Wat. Sci. Technol.
57(4), 575-579.
Clauwaert, P., Van Der Ha, D., Boon, N., Verbeken, K., Verhaege, M., Rabaey, K., Verstraete,
W., 2007b. Open air biocathode enables effective electricity generation with microbial fuel cells.
Environ. Sci. Technol. 41(21), 7564-7569.
Clauwaert, P., Verstraete, W., 2008. Methanogenesis in membraneless microbial electrolysis
cells. Appl. Microbiol. Biotechnol. 82(5), 829-836.
Clauwaert, P., Verstraete, W., 2009. Methanogenesis in membraneless microbial electrolysis
cells. Appl. Microbiol. Biotechnol. 82(5), 829-836.
Closed Reflux Colorimetric Method, 1992. Section 5220 D. Standard methods for the
examination of water and wastewater, 18 ed.
Cord-Ruwisch, R., Lovley, D.R., Schink, B., 1998. Growth of Geobacter sulfurreducens with
acetate in syntrophic cooperation with hydrogen-oxidizing anaerobic partners. Appl. Environ.
Microbiol. 64(6), 2232-2236.
Cord-Ruwisch, R., Mercz, T.I., Hoh, C.Y., Strong, G.E., 1997. Dissolved hydrogen
concentration as an on-line control parameter for the automated operation and optimization of
anaerobic digesters. Biotechnol. Bioeng. 56(6), 626-634.
Cord-Ruwisch, R., Seitz, H.-J., Conrad, R., 1988. The capacity of hydrogenotrophic anaerobic
bacteria to compete for traces of hydrogen depends on the redox potential of the terminal
electron acceptor. Arch. Microbiol 149, 350- 357.
Cortez, S., Teixeira, P., Oliveira, R., Mota, M., 2008. Rotating biological contactors: a review on
main factors affecting performance. Rev. Environ. Sci. Biotechnol. 7(2), 155-172.
Davis, J.B., Yarbrough, H.F., 1962. Preliminary experiments on a microbial fuel cell. Science
137, 615-616.
De Graef, M.R., Alexeeva, S., Snoep, J.L., De Mattos, M.J.T., 1999. The steady-state internal
redox state (NADH/NAD) reflects the external redox state and is correlated with catabolic
adaptation in Escherichia coli. J. Bacteriol. 181(8), 2351-2357.
References
- 172 -
de Mes, T.Z.D., Stams, A.J.M., Reith, J.H., Zeeman, G., 2003. Methane production by anaerobic
digestion of wastewater and solid wastes. In: Reith, J.H., Wijffels, R.H., Barten, H. (Eds.), Bio-
methane & Bio-hydrogen: Status and perspectives of biological methane and hydrogen
production, pp. 58-102. Dutch Biological Hydrogen Foundation, The Netherlands.
Di Palma, L., Verdone, N., 2009. The effect of disk rotational speed on oxygen transfer in
rotating biological contactors. Bioresourc. Technol. 100(3), 1467-1470.
Dumas, C., Basseguy, R., Bergel, A., 2008a. Electrochemical activity of Geobacter
sulfurreducens biofilms on stainless steel anodes. Electrochim. Acta 53, 5235-5241.
Dumas, C., Basseguy, R., Bergel, A., 2008b. Microbial electrocatalysis with Geobacter
sulfurreducens biofilm on stainless steel cathodes. Electrochim. Acta 53, 2494-2500.
EIA, 2008. International Energy Outlook 2008. Energy Information Administration, Official
Energy Statistics from the U.S. Government.
Fan, Y., Hu, H., Liu, H., 2007. Sustainable power generation in microbial fuel cells using
bicarbonate buffer and proton transfer mechanisms. Environ. Sci. Technol. 41(23), 8154-8158.
Finkelstein, D.A., Tender, L.M., Zeikus, J.G., 2006. Effect of electrode potential on electrode-
reducing microbiota. Environ. Sci. Technol. 40, 6990-6995.
Freguia, S., Rabaey, K., Yuan, Z., Keller, J., 2007a. Electron and carbon balances in microbial
fuel cells reveal temporary bacterial storage behavior during electricity generation. Environ. Sci.
Technol. 41(8), 2915-2921.
Freguia, S., Rabaey, K., Yuan, Z., Keller, J., 2007b. Non-catalyzed cathodic oxygen reduction at
graphite granules in microbial fuel cells. Electrochimica Acta 53(2), 598-603.
Freguia, S., Rabaey, K., Yuan, Z., Keller, J., 2007c. Sequential anode-cathode configuration
improves cathodic oxygen reduction and effluent quality of microbial fuel cells. Wat. Res. 42,
1387-1396.
Fricke, K., Harnisch, F., Schroder, U., 2008. On the use of cyclic voltammetry for the study of
anodic electron transfer in microbial fuel cells. Energy Environ. Sci. 1, 144-147.
Gil, G.C., Chang, I.S., Kim, B.H., Kim, M., Jang, J.K., Park, H.S., Kim, H.J., 2003. Operational
parameters affecting the performance of a mediator-less microbial fuel cell. Biosensors
Bioelectronics 18, 327-334.
Gregory, K.B., Bond, D.R., Lovley, D.R., 2004. Graphite electrodes as electron donors for
anaerobic respiration. Environ. Microbiol. 6 (6), 596-604.
Hallenbeck, P.C., 2005. Fundamentals of the fermentative production of hydrogen. Water Sci.
Technol. 52, 21-29.
References
- 173 -
Hallenbeck, P.C., Benemann, J.R., 2002. Biological hydrogen production; fundamentals and
limiting processes. Int. J. Hydrogen Energy 27 1185 - 1193.
Hallenbeck, P.C., Ghosh, D., 2009. Advances in fermentative biohydrogen production: the way
forward? Trends in Biotechnol. 27(5), 287-297.
Han, S.K., Shin, H.S., 2004. Biohydrogen production by anaerobic fermentation of food waste.
Int. J. Hydrogen Energy 29 569 - 577.
Harnisch, F., Schroder, U., 2009. Selectivity versus mobility: How to separate anode and
cathode in microbial bioelectrochemical systems? Chem. Sus. Chem. 2(10), 921-926.
Harnisch, F., Schroder, U., Scholz, F., 2008. The suitability of monopolar and bipolar ion
exchange membranes as separators for biological fuel cells. Environ. Sci. Technol. 42(5), 1740-
1746.
He, Z., Angenent, L.T., 2006. Application of bacterial biocathodes in microbial fuel cells.
Electroanalysis 18(19-20), 2009-2015.
He, Z., Minteer, S.D., Angenent, L.T., 2005. Electricity generation from artificial wastewater
using an upflow microbial fuel cell. Environ. Sci. Technol. 39, 5262-5267.
Heyndrickx, M., De Vos, P., De Ley, J., 1991. Fermentation characteristics of Clostridium
pasteurianum LMG 3285 grown on glucose and mannitol. J. Appl. Bacteriol. 70, 52-58.
Hoh, C.Y., 1996. Modelling of reaction kinetics in anaerobic ecosystems with considerations of
thermodynamic principles. School of Biological and Environmental Science. Murdoch
University, Perth, Western Australia.
Hu, H., Fan, Y., Liu, H., 2008. Hydrogen production using single-chamber membrane-free
microbial electrolysis cells. Wat. Res. 42(15), 4172-4178.
Hu, Z., 2008. Electrochemical determination of anaerobic microbial decay coefficients.
Chemosphere 72(2), 312-318.
IEA, 2004. World Energy Outlook 2004. International Energy Agency, Paris, France.
IPCC, 2007. Intergovernmental Panel on Climate Change. Climate Change 2007: Synthesis
Report. United Nations, New York.
Israni, S.H., Koli, S.S., Ratwardhan, A.W., Melo, J.S., D'souza, S.F., 2002. Phenol degradation
in rotating biological contactors. J. Chem. Technol. Biotechnol. 77(9), 1050-1057.
Ito, T., Nakashimada, Y., Senba, K., Matsui, T., Nishio, N., 2005. Hydrogen and ethanol
production from glycerol-containing wastes discharged after biodiesel manufacturing process. J.
Bioscience Bioeng. 100(3), 260-265.
References
- 174 -
Jackson, B.E., McInerney, M.J., 2002. Anaerobic microbial metabolism can proceed close to
thermodynamic limits. Nature 415, 454-456.
Jang, J.K., Pham, T.H., Chang, I.S., Kang, K.H., Moon, H., Cho, K.S., Kim, B.H., 2004.
Construction and operation of a novel mediator- and membrane-less microbial fuel cell. Process
Biochem. 39, 1007-1012.
Jeremiasse, A.W., Hamelers, H.V.M., Buisman, C.J.N., 2010. Microbial electrolysis cell with a
microbial biocathode. Bioelectrochemistry 78(1), 39-43.
Johansson, T.B., Goldemberg, J., 2004. World energy assessment overview: 2004 update.
United Nations Development Programme, UN-DESA and the World Energy Council.
Karube, I., Urano, N., Matsunaga, T., Suzuki, S., 1982. Hydrogen production from glucose by
immobilized growing cells of Clostrium butyricum. Eur. J. Appl. Microbiol. 16, 5-9.
Keighery, P., 2004. Risks in assessing the aeration requirements for an intermittent treatment
plant. Enviro 04 Convention and Exhibition, Sydney, Australia.
Kim, B.H., Chang, I.S., Gadd, G.M., 2007. Challenges in microbial fuel cell development and
operation. Appl Microbiol Biotechnol 76(3), 485-494.
Kim, B.H., Park, H.S., Kim, H.J., Kim, G.T., Chang, I.S., Lee, J., Phung, N.T., 2004a.
Enrichment of microbial community generating electricity using a fuel-cell-type electrochemical
cell. Appl. Microbiol. Biotechnol. 63, 672-681.
Kim, H.J., Park, H.S., Hyun, M.S., Chang, I.S., Kim, M., 2002. A mediator-less microbial fuel
cell using a metal reducing bacterium, Shewanella putrefaciens. Enzyme Microbial Technol. 30,
145-152.
Kim, S.H., Han, S.K., Shin, H.S., 2004b. Feasibility of biohydrogen production by anaerobic co-
digestion of food waste and sewage sludge. Int. J. Hydrogen Energy 29, 1607 - 1616.
Klass, D.L., 1993. Fossil fuel consumption and atmosphric CO2. Energy Policy 21(11), 1076-
1078.
Klass, D.L., 2003. A critical assessment of renewable energy usage in the USA. Energy Policy
31, 353-367.
Kumar, N., Das, D., 2000. Enhancement of hydrogen production by Enterobacter cloacae.
Process Biochem. 35, 589-593.
Lay, J.J., Lee, Y.J., Noike, T., 1999. Feasibility of biological hydrogen production from organic
faction of municipal solid waste. Water Res. 33, 2579-2586.
Lee, C.C., Lin, S.D., 2000. Handbook of Environmental Engineering Calculations, 1 ed.
McGraw-Hill Professional.
References
- 175 -
Lee, H.-S., Torres, C.I., Rittmann, B.E., 2009. Effects of substrate diffusion and anode potential
on kinetic parameters for anode-respiring bacteria. Environ. Sci. Technol. 42, 7571-7577.
Lee, J., Phung, N.T., Chang, I.S., Kim, B.H., Sung, H.C., 2006. Use of acetate for enrichment of
electrochemically active microorganisms and their 16S rDNA analyses. FEMS Microbiol. Lett.
223(2), 185-191.
Lehninger, A.L., Nelson, D.L., Cox, M.M., 1993. Principles of Biochemistry, 2 ed. Worth
Publishers.
Liu, H., 2002. Bio-hydrogen production from carbohydrate-containing wastewater. Department
of Civil Engineering, pp. 1-172. The University of Hong Kong, Hong Kong SAR.
Liu, H., Grot, S., Logan, B.E., 2005. Electrochemically assisted microbial production of
hydrogen from acetate. Environ. Sci. Technol. 39, 4317-4320.
Liu, Y., Harnisch, F., Fricke, K., Sietmann, R., Schroder, U., 2008. Improvement of the anodic
bioelectrocatalytic activity of mixed culture biofilms by a simple consecutive electrochemical
selection procedure. Biosensors. Bioelectronics 24, 1006-1011.
Logan, B., 2005a. Generating electricity from wastewater treatment. Water Environ. Res. , page
211 (editorial).
Logan, B., Regan, J.M., 2006a. Electricity-producing bacterial communities in microbial fuel
cells. Trends Microbiol. 14(12), 512-518.
Logan, B., Regan, J.M., 2006b. Microbial fuel cells: challenges and applications. Environ. Sci.
Technol. 40(18), 5172-5180.
Logan, B.E., 2005b. Simultaneous wastewater treatment and biological electricity generation.
Water Sci. Technol. 52(1-2), 31-37.
Logan, B.E., 2008. Microbial fuel cells. John Wiley & Sons, Inc., Hoboken, New Jersey.
Logan, B.E., Call, D., Cheng, S., Hamelers, H.V.M., Sleutels, T.H.J.A., Jeremiasse, A.W.,
Rozendal, R.A., 2008. Microbial electrolysis cells for high yield hydrogen gas production from
organic matter. Environ. Sci. Technol. 42(23), 8630-8640.
Logan, B.E., Hamelers, B., Rozendal, R.A., Schroder, U., Keller, J., Freguia, S., Aelterman, P.,
Verstraete, W., Rabaey, K., 2006. Microbial fuel cells: methodology and technology. Environ.
Sci. Technol. 40(17), 5181 -5192.
Lovley, D.R., 2006a. Bug juice: harvesting electricity with microorganisms. Nat. Rev. 4, 497-
508.
Lovley, D.R., 2006b. Microbial fuel cells: novel microbial physiologies and engineering
approaches. Curr. Opin. Biotechnol. 17, 327-332.
References
- 176 -
Lovley, D.R., 2008. The microbe electric: conversion of organic matter to electricity. Curr. Opin.
Biotechnol. 19(6), 564-571.
Lovley, D.R., Goodwin, S., 1988. Hydrogen concentrations as an indicator of the predominant
terminal electron-accepting reactions in aquatic sediments. Geochimica et Cosmochimica Acta
52, 2993-3003.
Lovley, D.R., Holmes, D.E., Nevin, K.P., 2004. Dissimilatory Fe(III) and Mn(IV) reduction Adv.
Microb. Physiol. 49 219-286.
Malki, M., De Lacey, A.L., Rodríguez, N., Amils, R., Fernandez, V.M., 2008. Preferential use
of an anode as an electron acceptor by an acidophilic bacterium in the presence of oxygen. Appl.
Environ. Microbiol. 74(14), 4472-4476.
Marcus, K.A., Torres, C.I., Rittmann, B.E., 2007. Conduction-based modelling of the biofilm
anode of a microbial fuel cell. Biotechnol. Bioeng. 98, 1171-1182.
Marken, F., Neudeck, A., Bond, A.M., 2002. Cyclic voltammetry. In: Scholz, F. (Ed.),
Electroanalytical methods: guide to experiments and applications, pp. 51-97. Springer-Verlag,
Berlin Heidelberg.
Mason, C.M., Culvern, J.B., 1949. Electrical conductivity of orthophosphoric acid and of
sodium and potassium dihydrogen phosphates at 25oC. J. Am. Chem. Soc. 71, 2387-2393.
Mench, M.M., Wang, C.-Y., Thynell, S.T., 2001. An introduction to fuel cells and related
transport phenomena. Int. J. Transport Phenomena 3, 151-176.
Menicucci, J., Beyenal, H., Marsili, E., A., V.R., Lewandowski, Z., 2006. Procedure for
determining maximum sustainable power generated by microbial fuel cells. Environ. Sci.
Technol. 40, 1062-1068.
Momirlan, M., Veziroglu, T.N., 2005. The properties of hydrogen as fuel tomorrow in
sustainable energy system for a cleaner planet. Int. J. Hydrogen Energy 30, 795-802.
Myers, C.R., Nealson, K.H., 1990. Respiration-linked proton translocation coupled to anaerobic
reduction of maganese(IV) and iron(III) in Shewanella putrefaciens MR-1. J. Bacteriol. 172,
6232-6238.
Nealson, K.H., Saffarini, D.A., 1994. Iron and manganese in anaerobic respiration:
environmental significance, physiology, and regulation. Annu. Rev. Microbiol. 48, 311-343.
Newman, D.K., Kolter, R., 2000. A role for excreted quinones in extracellular electron transfer.
Nature 405, 94-97.
Odom, J.M., Peck, H.D., 1984. Hydrogenase, electron-transfer proteins, and energy coupling in
the sulfate-reducing bacteria Desulfovibrio. Ann. Rev. Microbiol. 38, 551-592.
References
- 177 -
Oh, S.E., Kim, J.R., Joo, J.H., Logan, B.E., 2009. Effects of applied voltages and dissolved
oxygen on sustained power generation by microbial fuel cells. Wat. Sci. Technol. 60(5), 1311-
1317.
Oh, S.E., Min, B., Logan, B.E., 2004. Cathode performance as a factor in electricity generation
in microbial fuel cells. Environ. Sci. Technol. 38, 4900-4904.
Park, D.H., Kim, S.K., Shin, I.H., Jeong, Y.J., 2000. Electricity production in biofuel cell using
modified graphite electrode with neutral red. Biotechnol. Lett. 22, 1301-1304.
Park, D.H., Zeikus, J.G., 2000. Electricity generation in microbial fuel cells using neutral red as
an electronophore. Appl. Environ. Microbiol. 66(4), 1292-1297.
Parot, S., Delia, M.-L., Bergel, A., 2007. Forming electrochemically active biofilms from garden
compost under chronoamperometry. Bioresour. Technol. 99(11), 4809-4816.
Pham, H.T., Boon, N., Aelterman, P., Clauwaert, P., Schamphelaire, L.D., Vanhaecke, L.,
Maeyer, K.D., Hofte, M., Verstraete, W., Rabaey, K., 2007. Metabolites produced by
Pseudomonas sp. enable a Gram-positive bacterium to achieve extracellular electron transfer.
Appl. Microbiol. Biotechnol. DOI 10.1007/s00253-007-1248-6.
Pham, H.T., Rabaey, K., Aelterman, P., Clauwaert, P., De Schamphelaire, L., Verstraete, W.,
2006. Microbial fuel cells in relation to conventional anaerobic digestion technology. Eng. Life
Sci. 6(3), 285-292.
Pham, T.H., Aelterman, P., Verstraete, W., 2009. Bioanode performance in bioelectrochemical
systems: recent improvements and prospects. Trends Biotechnol. 27(3), 168-178.
Picioreanu, C., Loosdrecht, M.C.M., Cirtis, T.P., Scott, K., 2009. Model based evaluation of the
effect of pH and electrode geometry on microbial fuel cell performance. Bioelectrochemistry
doi:10.1016/j.bioelechem.2009.04.009.
Potter, M.C., 1911. Electrical effects accompanying the decomposition of organic compounds.
Proc. Roy. Soc. London, Series B. 84, 260-276.
Qiao, Y., Li, C.M., Bao, S.-J., Lu, Z., Hong, Y., 2008. Direct electrochemistry and
electrocatalytic mechanism of evolved Escherichia coli cells in microbial fuel cells. Chem.
Commun., 1290-1292.
Rabaey, K., Boon, N., Hofte, M., Verstraete, w., 2005a. Microbial phenazine production
enhances electron transfer in biofuel cells. Environ. Sci. Technol. 39, 3401-3408.
Rabaey, K., Boon, N., Siciliano, S.D., Verhaege, M., Verstraete, W., 2004. Biofuel cells select
for microbial consortia that self-mediate electron transfer. Appl. Environ. Microbiol. 70(9),
5373-5382.
References
- 178 -
Rabaey, K., Clauwaert, P., Aelterman, P., Verstraete, W., 2005b. Tubular microbial fuel cells for
efficient electricity generation. Environ. Sci. Technol. 39, 8077-8082.
Rabaey, K., Lissens, G., Siciliano, S.D., Verstraete, W., 2003. A microbial fuel cell capable of
converting glucose to electricity at high rate and efficiency. Biotechnol. Lett. 25, 1531-1535.
Rabaey, K., Lissens, G., Verstraete, w., 2005c. Microbial fuel cells: performances and
perspectives. In: Lens, P., Westermann, P., Haberbauer, M., Moreno, A. (Eds.), Biofuels for fuel
cells: renewable energy from biomass fermentation, pp. 375-396. IWA Publishing, London, UK.
Rabaey, K., Read, S.T., Clauwaert, P., Freguia, S., Bond, P.L., Blackall, L.L., Keller, J., 2008.
Cathodic oxygen reduction catalyzed by bacteria in microbial fuel cells. ISME J., 1-9.
Rabaey, K., Rodrı´guez, J., Blackall, L.L., Keller, J., Gross, P., Batstone, D., Verstraete, W.,
Nealson, K.H., 2007. Microbial ecology meets electrochemistry: electricity-driven and driving
communities. ISME J. 1, 9-18.
Rabaey, K., Verstraete, W., 2005. Microbial fuel cells: novel biotechnology for energy
generation. Trends Biotechnol. 23(6), 291-298.
Razi, F., Yassin, A.A.A., Lyuke, S.E., Ngan, M.A., Morimoto, M., 2005. Bio-hydrogen
systhesis from wastewater by anaerobic fermentation using microflora. Int. J. Green Energy 2,
387-396.
Reguera, G., D., M.K., Mehta, T., Nicoll, J.S., Tuominen, M.T., Lovley, D.R., 2005.
Extracellular electron transfer via microbial nanowires. Nature 435, 1098-1101.
Reguera, G., Nevin, K.P., Nicoll, J.S., Covalla, S.F., Woodard, T.L., Lovley, D.R., 2006.
Biofilm and nanowire production leads to increased current in Geobacter sulfurreducens fuel
cells. Appl. Environ. Microbiol. 72, 7345-7348.
Richter, H., Nevin, K.P., Jia, H., Lowy, D.A., Lovley, D.R., Tender, L.M., 2009. Cyclic
voltammetry of biofilms of wild type and mutant Geobacter sulfurreducens on fuel cell anodes
indicates possible roles of OmcB, OmcZ, type IV pili, and protons in extracellular electron
transfer. Energy Environ. Sci. 2, 506-516.
Rifkin, J., 2002. The Hydrogen Economy. Tarcher.
Rismani-Yazdi, H., Carver, S.M., Christy, A.D., Tuovinen, O.H., 2008. Cathodic limitation in
microbial fuel cells: an overview. J. Power Sources 180(2), 683-694.
Rittmann, B.E., 2008. Opportunities for renewable bioenergy using microorganisms. Biotechnol.
Bioeng. 100(2), 203-212.
Rozendal, R.A., 2007. Hydrogen production throught biocatalyzed electrolysis. Wageningen
University, Wageningen, the Netherlands.
References
- 179 -
Rozendal, R.A., Hamelers, H.V.M., Buisman, C.J.N., 2006a. Effects of membrane cation
transport on pH and microbial fuel cell performance. Environ. Sci. Technol. 40(17), 5206-5211.
Rozendal, R.A., Hamelers, H.V.M., Euverink, G.J.W., Metz, S.J., Buisman, C.J.N., 2006b.
Principle and perspectives of hydrogen production through biocatalyzed electrolysis. Int. J.
Hydrogen Energy 31, 1632-1640.
Rozendal, R.A., Hamelers, H.V.M., Rabaey, K., Keller, J., Buisman, C.J.N., 2008a. Towards
practical implementation of bioelectrochemical wastewater treatment. Trends Biotechnol. 26(8),
450-459.
Rozendal, R.A., Jeremiasse, A.W., Hamelers, H.V.M., Buisman, C.J.N., 2008b. Hydrogen
production with a microbial biocathode. Environ. Sci. Technol. 42(2), 629-634.
Ruzicka, M., 1996. The effect of hydrogen on acidogenic glucose cleavage. Water Res. 30(10),
2447-2451.
Schink, B., 1997. Energetics of syntrophic cooperation in methanogenic degradation. Microbiol.
Mol. Biol. Rev. 61, 262-280.
Schnurer, A., Schink, B., Svensson, B.H., 1996. Clostridium ultunense sp. nov., a mesophilic
bacterium oxidizing acetate in syntrophic association with a hydrogenotrophic methanogenic
bacterium. Int. J. Syst. Bacteriol. 46(4), 1145-1152.
Scholz, F., 2002. Electroanalytical methods: guide to experiments and applications. Springer-
Verlag Berlin Heidelberg.
Schroder, U., 2007. Anodic electron transfer mechanisms in microbial fuel cells and their energy
efficiency. Phys. Chem. Chem. Phys. 9, 2619-2629.
Schroder, U., 2008. From wastewater to hydrogen: biorefineries based on microbial fuel-cell
technology. Chem. Sus. Chem. 1(4), 281-282.
Sears, H.J., Spiro, S., Richardson, D.J., 1997. Effect of carbon substrate and aeration on nitrate
reduction and expression of the periplasmic and membrane-bound nitrate reductases in carbon-
limited continuous cultures of Paracoccus denitrificans Pd1222. Microbiology 143, 3767-3774.
Stams, A.J.M., de Bok, F.A.M., Plugge, C.M., van Eekert, M.H.A., Dolfing, J., Schraa, G., 2006.
Exocellular electron transfer in anaerobic microbial communities. Environ. Microbiol. 8(3), 371-
382.
Sund, C.J., McMasters, S., Crittenden, S.R., Harrell, L.E., Sumner, J.J., 2007. Effect of electron
mediators on current generation and fermentation in a microbial fuel cell. Appl Microbiol
Biotechnol 76(3), 561-568.
References
- 180 -
Svensson, B.H., Karlsson, A., 2005. Dark fermentation for hydrogen production from organic
wastes. In: Lens, P., Westermann, P., Haberbauer, M., Moreno, A. (Eds.), Biofuels for fuel cells:
renewable energy from biomass fermentation, pp. 209-220. IWA Publishing, London, UK.
Taguchi, F., Chang, J.D., Mizukami, N., Saito-Taki, T., Hasegawa, K., Morimoto, M., 1993.
Isolation of a hydrogen-producing bacterium, Clostridium beijerinckii strain AM21B, from
termites. Can. J. Microbiol. 39, 726-730.
Tanisho, S., 2001. A scheme for developing the yield of hydrogen by fermentation. In: Miyake,
J., Matsunnaga, T., San Pietro, A. (Eds.), Biohydrogen II, pp. 131-138. Elsevier Science,
Amsterdam.
Tanisho, S., Kamiya, N., Wakao, N., 1989. Microbial fuel cell using Enterobacter aerogenes.
Bioelectrochem. Bioenergetics 21, 25-32.
Tchobanoglous, G., Burton, F.L., Stensel, H.D., 2003. Wastewater engineering, treatment and
reuse, 4 ed. McGraw-Hill, Boston.
Thauer, R.K., Jungermann, K., Decker, K., 1977. Energy conservation in chemotrophic
anaerobic bacteria. Bacteriol. Rev. 41(1), 100-180.
Torres, C.I., Marcus, A.K., Parameswaran, P., Rittmann, B.E., 2008a. Kinetic experiments for
evaluating the Nernst-Monod model for anode-respiring bacteria (ARB) in a biofilm anode.
Environ. Sci. Technol. 42(17), 6593-6597.
Torres, C.I., Marcus, A.K., Rittmann, B.E., 2007. Kinetics of consumption of fermentation
products by anode-respiring bacteria. Appl. Microbiol. Biotechnol. DOI 10.1007/s00253-007-
1198-z.
Torres, C.I., Marcus, A.K., Rittmann, B.E., 2008b. Proton transport inside the biofilm limits
electrical current generation by anode-respiring bacteria. Biotechnol. Bioeng. 100, 872-881.
Van Aalst-van Leeuwen, M.A., Pot, M.A., Van Loosdrecht, M.C.M., 1997. Kinetic modelling of
poly(β-hydroxybutyrate) production and consumption by Paracoccus pantotrophus under
dynamic substrate supply. Biotechnol. & Bioeng. 55(5), 773-782.
van Ginkel, S.W., Logan, B., 2005. Increased biological hydrogen production with reduced
organic loading. Water Res. 39, 3819-3826.
van Lier, J.B., Mahmoud, N., Zeeman, G., 2008. Anaerobic wastewater treatment. In: Henze, M.,
van Loosdrecht, M.C.M., Ekama, G.A., Brdjanovic, D. (Eds.), Biological Wastewater Treatment:
Principles, Modelling and Design. IWA Publishing, London, UK.
van Lier, J.B., Tilche, A., Ahring, B.K., Macarie, H., Moletta, R., Dohanyos, M., Pol, L.W.H.,
Lens, P., Verstraete, W., 2001. New perspectives in anaerobic digestion. Water Sci. Technol. 43,
1-18.
References
- 181 -
Van Loosdrecht, M.C.M., Pot, M.A., Heijnen, J.J., 1997. Importance of bacterial storage
polymers in bioprocesses. Wat. Sci. Technol. 35, 41-47.
Van Spanning, R.J.M., de Boer, A.P.N., Reijnders, W.N.M., De Gier, J.-W.L., Delorme, C.O.,
Stouthamer, A.H., Westerhoff, H.V., Harms, N., van der Oost, J., 1995. Regulation of oxidative
phosphorylation: The flexible respiratory network of Paracoccus denitrificans. J. Bioenergetics
Biomembranes.
Verstraete, W., Van de Caveye, P., Diamantis, V., 2009. Maximum use of resources present in
domestic "used water". Bioresourc. Technol. 100, 5537-5545.
Verstraete, W., van Vaerenbergh, E., 1986. Aerobic activated sludge. In: Rehm, H.J., Reed, G.
(Eds.), Biotechnology, pp. 43-112. VCH, Weinheim, Germany.
von Sperling, M., de Lemos Chernicharo, C.A., 2005. Biological wastewater treatment in warm
climate regions. IWA Publishing, London, UK.
Vrije, T.d., Claassen, P.A.M., 2003. Dark hydrogen fermentations. In: Reith, J.H., Wijffels, R.H.,
Barten, H. (Eds.), Bio-methane & Bio-hydrogen: status and perspectives of biological methane
and hydrogen production, pp. 103-123. Dutch Biological Hydrogen Foundation, The Hague,
Netherlands.
Wang, B., Han, J.-I., 2008. A single chamber stackable microbial fuel cell with air cathode.
Biotechnol. Lett. 31(3), 1573-6776.
Wang, Y.F., Tsujimura, S., Cheng, S.S., Kano, K., 2007. Self-excreted mediator from
Escherichia coli K-12 for electron transfer to carbon electrodes. Appl. Microbiol. Biotechnol. 76,
1439-1446.
Watanabe, K., 2008. Recent developments in microbial fuel cell technologies for sustainable
bioenergy. J. Biosci. Bioeng. 106(6), 528-536.
Yoon, S.-M., Choi, C.-H., Kim, M., Hyun, M.S., Shin, S.-H., Yi, D.-H., Kim, H.J., 2007.
Enrichment of electrochemically active bacteria using a three-electrode electrochemical cell. J.
Microbiol. Biotechnol. 17(1), 110-115.
You, S., Zhao, Q., Zhang, J., Jiang, J., Wan, C., Du, M., Zhao, S., 2007. A graphite-granule
membrane-less tubular air-cathode microbial fuel cell for power generation under continuously
operational conditions. J. Power Sources 173(1), 172-177.
You, S., Zhao, Q., Zhang, J., Jiang, J., Zhao, S., 2006. A microbial fuel cell using permanganate
as the cathodic electron acceptor. J. Power Sources 162, 1409-1415.
Yu, E.H., Cheng, S., Scott, K., Logan, B., 2007. Microbial fuel cell performance with non-Pt
cathode catalysts. J. Power Sources 171, 275-281.
References
- 182 -
Zhang, T., Liu, H., Fang, H.H.P., 2003. Biohydrogen production from starch in wastewater
under thermophilic condition. J. Environ. Manage. 69, 149-156.
Zhao, F., Harnisch, F., Schroder, U., Scholz, F., Bogdanoff, P., Herrmann, I., 2005. Application
of pyrolysed iron (II) phthalocyanine and CoTMPP based oxygen reduction catalysts as cathode
materials in microbial fuel cells. Electrochem. Commun. 7, 1405-1410.
Zhao, F., Harnisch, F., Schroder, U., Scholz, F., Bogdanoff, P., Herrmann, I., 2006a. Challenges
and constraints of using oxygen cathodes in microbial fuel cells. Environ. Sci. Technol. 40(17),
5193-5199.
Zhao, F., Harnisch, F., Schroeder, U., Scholz, F., Bogdanoff, P., Herrmann, I., 2006b.
Challenges and constraints of using oxygen cathodes in microbial fuel cells. Environ. Sci.
Technol. 40(17), 5193-5199.
- 183 -
A Appendixes
Appendix 1 (Supplementary Note to Chapter 1, Section 1.1.2.2.2)
Limitations of Fermentative Hydrogen Production: An Overview
(This article has been published in American Institute of Physics Conf. Proc. 941, 264-269)
Abstract
Turning organic wastes into hydrogen (H2) by using anaerobic fermentative technology can
accomplish both waste treatment and energy recovery objectives. However, H2 production using
anaerobic fermentative approaches faces a fundamental limitation of poor H2 production yield
(i.e. < 4 mol H2 per mol of hexose). This article gives an overview on the limitations of bio-H2
production using fermentative processes. Fundamental microbiology and thermodynamic of the
processes are discussed.
1. Introduction
By decoupling H2 production from methane production, conversion of organic matter
into H2 could be achieved (Vrije and Claassen 2003). Many types of organic compounds,
ranging from polymers to monomers such as carbohydrates, fats and amino acids are known to
be the substrates for H2 production (Claassen et al. 1999). As such, organic wastes used for the
production of methane can also be the potential substrates for anaerobic fermentative production
of H2 (Svensson and Karlsson 2005). In fact, many studies indicate that various wastes
containing high organic matter have been used to produce H2 by anaerobic fermentation process
Appendix 1
- 184 -
(Angenent et al. 2004; Han and Shin 2004; Kim et al. 2004b; Lay et al. 1999; Razi et al. 2005;
van Ginkel and Logan 2005; Zhang et al. 2003). The major differences between the two
processes are that successful biological H2 production requires inhibition of H2-consuming
microorganisms and maximization of H2 production yield from the organic substrates. Most of
the studies revealed that only 10-20% of stoichiometric maximum yield (i.e. 12 moles of H2 per
mole hexose) of H2 could be recovered in anaerobic fermentative processes (Benemann 1996;
Hallenbeck 2005; Hallenbeck and Benemann 2002). Hence, some researchers even argued that
fermentative H2 production should only be restricted to a pre-treatment step in a larger bioenergy
production concept (Angenent et al. 2004). The present article will discuss, from a
microbiological and thermodynamic perspective, the limitations of anaerobic fermentation as H2
production process.
2. Biochemical Pathways of Fermentative Hydrogen Production
Among the many species of anaerobic fermentative bacteria capable of producing H2, the
H2-producing characteristics of two genera, Clostridium (e.g. C. pasteurianum (Heyndrickx et al.
1991); C. beijerinckii (Taguchi et al. 1993) and C. butyricum (Karube et al. 1982) and
Enterobacter (e.g. E. aerogenes (Tanisho et al. 1989) and E. Cloacae (Kumar and Das 2000),
have been studied extensively. Figure 1 illustrates the pathway of anaerobic fermentation by
using glucose as model substrate (Tanisho 2001). The glucose monomer produced from
hydrolysis is taken up by the fermentative bacteria and degraded predominantly through the
Embden-Meyerhof-Parnas (glycolysis) pathways to generate ATP from ADP, leading to the
formation of pyruvate (CH3COCOOH). Like other bacteria or higher eukaryotic cells, the
glycolytic reactions in glucose fermentative bacteria also produce electron equivalents in the
form of NADH, which has to be re-oxidized in order to continue substrate degradation.
Inorganic electron acceptors are usually the preferred candidate in anaerobic respiration for the
bacteria to regenerate these reducing equivalents because they have higher redox potential.
While in the absence of external inorganic electron acceptors, NADH is commonly reoxidized
by H+ and produce H2 and NAD
+. The pyruvate produced from glycolysis is then converted to
acetyl-CoA (CH3COSCoA), liberating carbon dioxide and H2. The pyruvate may also be
converted into acetyl-CoA and formate (CHOOH), which may be converted to H2 and carbon
dioxide by fermentative bacteria such as Escherichia coli. Depending on the microorganisms
Appendix 1
- 185 -
and the environment, the Acetyl-CoA is eventually converted into acetate (CH3COOH), butyrate
(CH3CH2CH2COOH) or ethanol (CH3CH2OH).
Figure A1.1. Pathways of fermentation of glucose. Adapted from (Liu 2002).
The formation of different end products from the pyruvate is found to be highly
dependent on the H2 partial pressure of the system. For instance, since the conversion of glucose
to butyrate, CO2 and H2 yield only 3 mol of ATP per mol of glucose, whereas the conversion of
glucose to acetate, CO2 and H2 can yield 4 mol of ATP per mol of glucose. The production of H2
and acetate from pyruvate is therefore usually preferred by most bacteria with hydrogenase as it
allows the bacteria to conserve more energy from their substrates (Thauer et al. 1977). In fact,
from bio-H2 production stand point, formation of acetate as the end product is also preferred
because it allows the greatest possible amount of H2 to be produced via fermentative pathway.
However, such preferred conversion is only possible when the H2 concentration (or
NADH/NAD+ ratio) in the system is kept low. Under elevated H2 concentration environment,
which may be due to accumulation of H2 because of ineffective H2 removal (e.g. inhibition of
H2-consuming methanogen), H2 production from NADH can be inhibited due to thermodynamic
C6H12O6
2 ADP
2 ATP
2 NAD+
2 NADH + 2 H+
2 H2
2 CH3COCOOH
2 CH3COSCoA
CH3CH2CH2COOH
2 CH3COOH2 CH3CH2OH
4 NAD+ 4 NADH + 4 H+ ADP ATP
2 CO22 H2
2 CHOOH
2 H2
2 CO2
ADP
ATP 2 NAD+
2 NADH + 2 H+
Appendix 1
- 186 -
reasons, leading to the production of propionate or butyrate from pyruvate as alternative electron
sinks (Figure A1.2).
3. Elimination of Syntrophic Partners in Fermentative Hydrogen Production
In methanogensis process, complex organic matter is degraded in a sequence of reactions
by several distinct groups of microorganisms. Fermentative microorganisms will first
breakdown the organic matter into simpler substance, such as H2, formate and acetate, which
serve as the substrates for the methanogenic microorganisms. Also, a variety of other organic
compounds like lactate, ethanol, propionate, butyrate, etc. are formed, and are degraded by
proton-reducing acetogenic bacteria to form the methanogenic substrates, i.e. H2 and acetate.
However, these organisms can only grow and keep converting the organic matter into the
methanogenic substrates when the H2 concentration is maintained at low level by their
syntrophic partner, i.e. methanogenic archaea (Bryant et al. 1967; Schink 1997; Schnurer et al.
1996). Such H2 (or electron + proton) transfer mechanism is an integral and vital process in the
anaerobic mineralization of organic matter, and is commonly regarded as interspecies
interactions between fermentative H2-producing organisms and H2-consuming methanogen
(Odom and Peck 1984).
From a bioenergetic perspective, when organic electron acceptors such as pyruvate or
fumurate were used, the partial oxidation of the substrates results in the formation of products
that still possess high free energy. This explains why anaerobic catabolic reactions involving
organic electron acceptors usually have low free energy change (∆G), and only little energy
could be conserved from the substrate metabolism by the microorganisms (Thauer et al. 1977).
Therefore, the reactions proceed very close to the dynamic equilibrium, and require constant
removal of the end-products of the reactions before the reactions can proceed further (Jackson
and McInerney 2002; Schink 1997). Therefore, the presence of methanogenic organisms is vital
because they can effectively and indirectly maintain the continuous fermentative degradation of
Appendix 1
- 187 -
At Hydrogen Partial
Pressure lowers than
10 Pa:
At Hydrogen Partial
Pressure higher than
10 Pa:
Figure A1.2. Formation of acetate, propionate or butyrate via different pathways of
glucose/pyruvate degradation at elevated or low hydrogen concentrations. H2 and acetate
are produced at low H2 concentration (< 10 Pa) whilst propionate and butyrate are
produced as alternative electron sinks at elevated H2 due to thermodynamic reasons.
Adapted from (Hoh 1996).
organic matter into their own substrates, including H2. However, in fermentative H2 production
process, growth of methanogens is considered as an undesirable process and should be inhibited
Glucose
Pyruvate Pyruvate
ATP ATP
NADH NADH
ATP ATP
Acetate AcetateH2H2
H2H2
CO2 CO2
Glucose
Pyruvate Pyruvate
ATP ATP
NADH NADH
ATP ATP
Acetate AcetateH2H2H2H2
H2H2H2H2
CO2 CO2
Glucose
Pyruvate Pyruvate
ATP ATP
NADH NADH
ATP
Acetate Propionate
H2
CO2
Glucose
Pyruvate Pyruvate
ATP ATP
NADH NADH
ATP
Acetate Propionate
H2H2
CO2
Glucose
Pyruvate Pyruvate
ATP ATP
NADH NADH
ATP
Acetate
Butyrate
H2
CO2
H2
CO2
Glucose
Pyruvate Pyruvate
ATP ATP
NADH NADH
ATP
Acetate
Butyrate
H2H2
CO2
H2H2
CO2
Appendix 1
- 188 -
in order to prevent any consumption of H2, allowing the H2 produced from fermentative
degradation of organic matter by fermentative or acetogenic bacteria to be recovered.
4. Conversion of Fermentation End-Products into Hydrogen: A Thermodynamic
Consideration
A variety of fermentative end-products (mainly as VFAs) are remained after the
fermentative process (refer to Figure A1.1). However, a considerable amount of chemical energy
is still remaining ―unexplored‖ from the process. In methanogenesis, degradation of VFAs such
as acetate is only possible via interspecies H2 transfer, when H2 is kept to a low concentration by
the H2-consuming methanogens, accumulation of H2 to a certain level can result in the inhibition
of VFA degradation reactions due to thermodynamic limitations (Thauer et al. 1977). This
explains why dissolved H2 concentration is considered as a key controlling factor in the
anaerobic processes and it can regulate the rates at which the VFAs degrade to acetate and H2
(Cord-Ruwisch et al. 1997; Lovley and Goodwin 1988). The standard Gibbs free energy change
of reaction (∆Go) is the amount of energy being conserved when the substrate(s) are being
converted into products (s) (Thauer et al. 1977). It can be calculated from the Gibbs free energy
changes of formation from the elements (∆Go
f) of the substrates and products according to the
following equation:
∆Go =
∑
∆G
of of products - ∑
∆G
of of substrates
Since almost all reactions occur in biological systems are not under standard condition (i.e. 298
K, 1 atm pressure, 1 M concentration for all species). The Gibbs free energy change of reaction
corrected for the actual reacting conditions (∆G) can then be calculated from the ∆Go and the
concentrations of all reactants (S) and products (P) according to the following equation:
∆G = ∆G
o + RT log [P]/[S]
Where, R = universal gas constant (8.314 x 10-3
kJ∙mol-1
∙K-1
); T = absolute temperature (K).
(Negative values of ∆G indicate the forward reaction is exergonic or spontaneous,
Appendix 1
- 189 -
and positive values of ∆G indicate the backward reaction is exergonic or
spontaneous.)
According to the literature, amongst all of the end-products commonly formed during
fermentative H2 production, acetate is considered as the key end-product because further
conversion of acetate into H2 is unfavorable to occur in fermentative processes. Based on Gibbs
free energy calculations, H2 production from acetate requires a net energy input of 104.6 kJ mol
acetate-1
under standard conditions (Thauer et al. 1977).
Figure A1.3. Relationship between partial pressure of hydrogen (PH2) and change-in-free-
energy (∆G) for the following hydrogen-producing reactions:
∆Go‘
(kJ/reaction) PH2, ∆G=0
(A) Acetate- + 4H2O → 2HCO3
- + 4H2 + H
+ +104.6 -4.60
(B) Propionate- + 3H2O → acetate
- + 1HCO3
- + 3H2 + H
+ +76.5 -4.49
(C) Butyrate- + 2H2O → acetate
- + 2H2 + H
+ +48.3 -4.25
(D) Ethanol + H2O → acetate- + 2H2 + H
+ +9.6 -0.85
(E) Glycerol + 2H2O → acetate- + HCO3
- + 3H2 + 2H
+ -84.6 +4.96
-10
-5
0
5
10
15
-100-80-60-40-20020406080100120
-10
-5
0
5
10
15
-100-80-60-40-20020406080100120
PH2
(Log10 atm)
∆G at pH 7, 25oC
(A)
(B)
(C)
(D)
(F)
(E)
Appendix 1
- 190 -
(F) Glucose + 4H2O → 2 acetate- + 2HCO3
- + 4H2 + 4H
+ -206.1 +9.06
Notes: PH2, ∆G=0 = hydrogen partial pressure (Log10 atm) when ∆G is equal to zero; ∆G
values (at pH 7 and 25oC) were calculated from the standard Gibbs free energy change (∆G
o‘)
value in (Thauer et al. 1977) using equation ∆G = ∆Go‘
+ 2.47 x ln ([products]/[substrates]).
Concentrations used for all compounds were assumed to be 1M.
Figure A1.3 illustrates the relationship between H2 partial pressure and change-in-free-
energy for several H2 producing reactions that possibly occur in fermentative process. Under
standard condition (i.e. 298 K, 1 atm pressure, 1 M concentration for all species), only the
oxidation reactions of 1 mol glucose to 4 mol H2 and 1 mol glycerol to 3 mol H2 have negative
∆Go‘
values, while all the other reactions have positive ∆Go‘
values, meaning that except for
glucose and glycerol, all the other reactions are energetically unfavorable to occur under
standard condition.
Since ∆G value of each specific reaction is in a linear function of partial pressure of the
product H2, the reaction may proceed toward the product side when the partial pressure of H2
has reduced to a certain level. When ∆G is equal to zero, oxidation of 1 mol acetate to 4 mol H2
only become energetically favorable when the partial pressure of H2 in the system is lower than -
4.6 Log10 atm (i.e. 2.55 Pa), which is extremely low that may even not allow the synthrophic
action of the hydrogenotrophic methanogenic bacteria in anaerobic fermentative system to occur.
According to Figure A1.3, glycerol should also be considered as a promising feedstock to
produce H2 via fermentative processes. This is because the conversion of glycerol into H2 is
energetically favorable (i.e. ∆Go‘
= -84.6 kJ/reaction), and the reaction is only susceptible to the
inhibitory effect of H2 accumulation with H2 partial pressure greater than +4.96 Log10 atm (i.e.
9.24 x 109 Pa), which is very high and is not supposed to happen under biological conditions. In
fact, glycerol is the major by-product of bio-diesel generation processes, and a further increase
in the production of bio-diesel fuels would raise the problem of efficiently treating wastes
containing glycerol (Ito et al. 2005). Therefore, apart from acetate, the possibility of using
glycerol as the feedstock substrates in H2 production process also worthwhile to be investigated.
As aforementioned, even the fermentative process can be successfully manipulated to use
acetate as the substrate for the production of H2, the produced H2 would be under very low
partial pressure, i.e. concentration. This is impractical to obtain large quantity of H2 using
Appendix 1
- 191 -
acetate as the only substrate from fermentative process. In order for the microbes to ―extract‖ the
electrons from the substrate in a more effective manner, it is reasonable to think of using some
approaches to remove the H2 that is produced from the bacteria immediately so as to maintain a
low H2 partial pressure for the H2 producing reactions to remain energetically favorable.
In methanogenesis process where methane forming consortia are present, effective H2
removal is normally being achieved by the methanogenic partners, which are in syntrophic
relationship at close proximity with the H2-producing bacteria. As H2 is one of the substrates for
methanogens, H2 is taken up by them immediately and effectively to produce methane, which
will then be easily collected in gaseous form due to their poor aqueous solubility. In other words,
the distance between the locations of H2 production and H2 consumption is so close to enable
effective mass transfer of H2 between the two partners. In fact, effective mass transfer of H2
between H2-producing and –consuming bacteria can be revealed by estimating the time required
for H2 removal from H2-producing bacteria in a methanogenic environment where H2 is
maintained at low level by syntrophic microbial association.
Time required for the H2-producing bacteria to produce H2 until Inhibitory Levels are reached
under Methanogenic Conditions:
Assume: Methane production rate = 1 L-1
CH4 L-1
day-1
Since: 1 mol of methane formed required 4 mol of H2, i.e. 4H2 + CO2 ↔ CH4 + 2H2O; and
molar volume of CH4 and H2 are 24.5 L mol-1
at 25 oC, 1 atm.
Then: H2 production rate = 4 L-1
H2 L-1
day-1
= 0.16 mol H2 L-1
day-1
= 0.16 M day-1
= 1.85 x 10-6
M second-1
= 1.85 µM second-1
Assume: Saturation concentration of H2 in water (25 oC,1 atm) is 1 mmol L
-1 (Ruzicka 1996)
Appendix 1
- 192 -
If: H2 production was inhibited at PH2 of 10 Pa (Cord-Ruwisch et al. 1998); i.e. 1 x 10-4
mmol L-1
= 100 nM = 0.1 µM
Then: Time Required to Produce H2 until Inhibitory Levels are reached would be:
= 0.1 µM/ 1.85 µM second-1
= 54 milliseconds
This simple calculation suggests that the methanogenic partner (H2-consuming
microorganisms) could effectively take up the H2 produced by the H2-producing bacteria within
a very short time in the system (i.e. approx. <0.05 second!). Hence, in order to allow H2 to be
collected from the process, it is an important yet challenging task to avoid H2 consumption by
the H2-consuming methanogen in the system (by inhibiting the activity of methanogen) while at
the same time maintaining a low H2 partial pressure in order to remove the thermodynamic
bottleneck of the H2 producing reaction.
5. Conclusion
Undoubtedly, anaerobic fermentation of organic substrates would allow energy recovery
in the form of H2 gas. However, in terms of conversion efficiency the process releases only
about 17% of H2 in the substrate (i.e. glucose). Most of the H2 still remain in the end products
which resist for further H2 conversion due to microbiological and thermodynamical limitations.
Appendix 2
- 193 -
Appendix 2 (Supplementary information to Chapter 2, Section 2.3.1)
Estimation of Internal Resistance of the MFC
In Chapter 2 Table 1, the internal resistance (Rint) values are obtained by using the
polarization slope method (Logan 2008).
In the literature, there are several well accepted methods for the determination of Rint in a
MFC. These methods include electrochemical impedance spectroscopy (EIS) based on Nyquist
plots; current interrupt methods; power density peak and polarization slope method. The first
two methods require the use of a potentiostat that is able to varying the frequency of the
sinusoidal signal over a wide frequency range (typically from 10-4
to 106 Hz). The circuit
interruption method also requires the use of a potentiostat. This method also requires a very
effective and fast voltage measurement after interrupting the current.
Figure A2.1. Characteristics of a MFC polarization curve showing regions different types of
losses (overpotentials). In Region 2, the constant linear voltage drop allows an estimation
of internal resistance (Rint) of the MFC. Eo‘
cell = maximal electromotive force of the
redox reaction couples.
Due to the unavailability of suitable instruments for EIS and current interruption
methods, a relatively simple method, the polarization slope method was used to estimate the Rint
Current
Cell
Volta
ge
Eo’cell
Cell Voltage = OCV* - IRint
Re
gio
n 1
: A
cti
va
tio
n lo
ss
es
Re
gio
n 2
: O
hm
iclo
ss
es
Re
gio
n 3
: C
on
ce
ntr
ati
on
los
ses
Open Circuit Voltage (OCV)
Appendix 2
- 194 -
of the described MFC in Chapter 2. The Rint of the MFC can be directly obtained from the slope
of the linear portion of the polarization curve. As depicted Figure A2.1, region 2. The Rint values
of the MFC recorded at different times during the acclimation period are shown in Figure A2.2.
Figure A2.2. Determination of MFC internal resistances at different times during the
acclimation period using the polarization slope method. The slopes of the linear
regression curves are the internal resistance value (Ω) of the respective system. Data
obtained from Figure 2.3.
y = -2.26x + 761.21
R2 = 1.00
0
200
400
600
800
0 50 100
y = -1.20x + 733.12
R2 = 1.00
0
200
400
600
800
0 50 100 150 200
y = -0.97x + 711.45
R2 = 0.99
0
200
400
600
800
0 50 100 150 200
y = -0.54x + 696.85
R2 = 0.99
0
200
400
600
800
0 50 100 150 200 250 300 350 400 450 500
Current (mA)
Cell
Voltage (
mV
)
Day 5 Day 14
Day 22 Day 30
Appendix 3
- 195 -
Appendix 3 (Supplementary information to Chapter 4, Section 4.3.2)
In Chapter 4, application of computer feedback in BES processes is introduced with an
example of running a cyclic voltammetry analysis by feedback controlling the external
resistance of a MFC. Yet, this is only an example of using the proposed method for evaluating
electrochemical behavior of an anodophilic biofilm in a MFC. Other practical applications could
be possible based on a similar concept. In this appendix, another practical example is given:
MFC are designed to generate electrical power. To operate a MFC at its maximal power
generating capacity requires the knowledge of the optimum (or suitable) external resistance (of
the power consumer)(Aelterman et al. 2008). This knowledge can be obtained by using
computer feedback control of the external resistance as well. However, instead of a cyclic
potentiostatic control of the electrode potential as described in Chapter 4, a steady state control
of different anodic potential set points would allow the MFC to exhibit steady state maximal
power output (here ranged from 225 to 250 W m-3
) and the corresponding external resistance
setting (here ranged from 7 to 8 Ω) (Figure A3.1 A, C).
With this approach, one could operate the MFC at this selected external resistance to
sustain a maximal power output of the MFC by using computer feedback. In summary, using
computer feedback for BES process control should be given attention in future development of
the technology.
Appendix 3
- 196 -
Figure A3.1. Potentiostatic control of different anodic potential set points by feedback
controlling the external resistance of a MFC. (A) anodic potential and external resistance
vs. time; (B) current-time plot; (C) power density-time plot; (D) cathodic potential and
electrolyte Eh vs. time. Notes: Selected anodic set points: -400, -380, -350, -300, -275, -
250, -200, -150, -100, -50 mV vs. Ag/AgCl). Anolyte was saturated with the electron
donor substrate (here about 10 mM acetate) throughout the experiment. Red dotted
arrows indicate maximal power output and the corresponding anodic potential (here
between -350 and -375 mV vs. Ag/AgCl).
0
20
40
60
80
100
Curr
ent
(mA
)
-600
-450
-300
-150
0
150
0 100 200 300 400Time (min)
Po
ten
tia
l
(mV
vs.
Ag
/Ag
Cl)
.
CatholyteCathodeAnolyte
-450
-400
-350
-300
-250
-200
-150
-100
-50
An
od
ic P
ote
ntia
l
(mV
vs. A
g/A
gC
l).
1
10
100
1000
Exte
rna
l Re
sis
tan
ce
(Oh
m) .
Anodic Potential (Set point)
External resistance
0
50
100
150
200
250
Po
we
r D
en
sity (
W m
-3)
A.
B.
C.
D.
- 197 -
– END of the Thesis –
BES is a very exciting scientific research topic. However, towards a real-world
“Waste-to-Energy” application of BES there is still a long way to go!
We are just at the beginning of this long journey…
2010, Ka Yu
CV - K.Y.CHENG
- 198 -
Ka Yu CHENG
— Contributing to the sustainable development of humankind using innovative approaches —
Education
April 2006- 2009 Ph.D. (Candidate) in Environmental Engineering
Faculty of Sustainability, Environmental and Life Sciences
Murdoch University, Perth, Australia
Jul 2003 – Jul 2005 M.Phil. in Environmental Science
Research Topic: Phyto-remediation of PAH contaminated soil
Department of Biology, Faculty of Science
Hong Kong Baptist University (HKBU)
Sep 2000 – Sep 2003 B.Sc (Honor) in Applied Biology (Environmental Science)
Research Topic: Partitioning of PAHs in soil-water systems
Department of Biology, Faculty of Science
Hong Kong Baptist University (HKBU)
Awards/Achievements
2008 Huber Technology Prize (1st Prize)
The prize honors ideas, concepts and results of research that are an innovative contribution to
the reuse of energy and valuable materials from wastewater. (In IFAT 2008 - 15th International
Trade Fair for Water - Sewage - Refuse - Recycling, Munich, Germany)
2005 PhD Scholarship Awarded:
1. Murdoch International Fee Waiver Scholarship (Murdoch U) (Accepted)
2. Murdoch University Research Scholarship (Murdoch U) (Accepted)
3. Monash International Postgraduate Research Scholarship (Monash U)
4. Monash Graduate Scholarship (Monash U)
5. Griffith University Postgraduate Research Scholarship (Griffith U)
6. Australian Government International Postgraduate Research Scholarship (Griffith U)
2005 The Best Poster Award (1st Prize), International Symposium on Phytoremediation and
Ecosystem Health 2005 at Hangzhou, China
2003 President’s Honor Roll (HKBU)
2001 Ambassador, Beijing Environmental Protection Study Visit 2001, Environmental Awards
Scheme for University Students 2001
Peer-Review Activity
2010 – Present Reviewer for journals Electrochimica Acta, Water Research, Environmental Science
and Technology
2009 – Present Reviewer for journal Bioelectrochemistry
Patent
K. Y. Cheng, R. Cord-Ruwisch, Goen Ho (2009). A Wastewater Treatment Process. Australian
Provisional Patent.
CV - K.Y.CHENG
- 199 -
Publications
K. Y. Cheng, G. Ho and R. Cord-Ruwisch. (2010) An anodophilic biofilm catalyzes cathodic oxygen
reduction. Environmental Science and Technology. Vol. 44(1), 518-525.
K. Y. Cheng, R. Cord-Ruwisch and G. Ho. (2009) A new approach for in situ cyclic voltammetry of a
microbial fuel cell biofilm without using a potentiostat. Bioelectrochemistry. Vol. 74, 227-231
K. Y. Cheng, G. Ho and R. Cord-Ruwisch (2008) Affinity of microbial fuel cell biofilm for the anodic
potential. Environmental Science and Technology. Vol. 42(10), 3828-3834.
K. Y. Cheng, K. M. Lai, J. W. C. Wong. (2008) Effects of pig manure compost and nonionic-surfactant
Tween 80 on phenanthrene and pyrene removal in soil vegetated with Agropyron elongatum.
Chemosphere. Vol. 73, 791-797.
K. Y. Cheng, J. W. C. Wong. (2008). Fate of 14C-pyrene in soil-plant system amended with pig manure
compost and Tween 80: a growth chamber study. Bioresource Technology. Vol. 99, 8406-8412.
K. Y. Cheng, R. Cord-Ruwisch and G. Ho. (2007) Limitations of bio-hydrogen production by anaerobic
fermentation process: an overview. American Institute of Physics Conf. Proc. Vol. 941, 264-269.
K. Y. Cheng, J. W. C. Wong. (2006). Effect of synthetic surfactants on the solubilization and distribution
of PAHs in water/soil-water systems. Environmental Technology. Vol. 27, 835-844.
K. Y. Cheng, J. W. C. Wong. (2006). Combined effect of nonionic surfactant Tween 80 and DOM on the
behavior of PAHs in soil-water system. Chemosphere. Vol. 62, 1907-1916.
K. Y. Cheng, Z. Y. Zhao, J. W. C. Wong. (2004). Solubilization and desorption of PAHs in soil-aqueous
system by biosurfactants produced from Pseudomonas aeruginosa P-CG3 under thermophilic
condition. Environmental Technology. Vol. 25, 1159-1165.
K. Y. Cheng, Z. Y. Zhao, J. W. C. Wong. (2004). Effects of biosurfactants produced from Pseudomonas
aeruginosa P-CG3 on the solubilization and desorption of PAHs in soil-aqueous system under
thermophilic condition. Proceedings of the Fourth International Conference on Remediation of
Chlorinated and Recalcitrant Compounds, Monterey, California, USA; Publisher: Battelle Press,
Columbus, Ohio.
Conference Presentations
2008 K. Y. Cheng, R. Cord-Ruwisch and G. Ho. A mixed anodophilic biofilm exhibits saturation
behavior with anodic potential in a microbial fuel cell. Microbial Fuel Cells First International
Symposium, Pennsylvania State University, 27-29 May 2008 at Pennsylvania State, USA (Poster)
2007 K. Y. Cheng, R. Cord-Ruwisch and G. Ho. Computer-controlled microbial fuel cell enables
efficient electricity production from activated sludge. IWA Specialist Conference: 11th World
Congress on Anaerobic Digestion: Bioenergy for Our Future – Renewable Energy from Waste.
23-27 Sep 2007 at Brisbane, Queensland, Australia (Poster)
2007 K. Y. Cheng, R. Cord-Ruwisch and G. Ho. Feasibility of bio-hydrogen production by anaerobic
fermentation process: A critical review. WREN International Conference Renewable Energy for
Sustainable Development in the Asia-Pacific Region. 4-8 Feb 2007 at Frementle, Perth, WA,
Australia (Oral)
2005 K. Y. Cheng and J. W. C. Wong. Phytoremediation of PAH contaminated soil using pig manure
compost and nonionic surfactant Tween 80. International Symposium on Phytoremediation and
Ecosystem Health. 10-13 Sep 2005 at Hangzhou, China (Best Poster Award)
2005 K. Y. Cheng and J. W. C. Wong*. Fate of 14C-pyrene in Soil-plant System Amended with Pig
Manure Compost and Tween 80: a Growth Chamber Study. International Symposium on
Phytoremediation and Ecosystem Health. 10-13 Sep 2005 at Hangzhou, China (*Oral)
2004 K. Y. Cheng, Z. Y. Zhao, J. W. C. Wong. Effects of biosurfactants produced from
Pseudomonas aeruginosa P-CG3 on the solubilization and desorption of PAHs in soil-aqueous
CV - K.Y.CHENG
- 200 -
system under thermophilic condition. Fourth International Conference on Remediation of
Chlorinated and Recalcitrant Compounds. 24-27 May 2004 at Monterey, California, US (Poster)